{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.440922190201729,
  "eval_steps": 3000,
  "global_step": 15000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.905435800552368,
      "epoch": 0.0004803073967339097,
      "grad_norm": 15.375,
      "learning_rate": 2e-06,
      "loss": 14.4544,
      "mean_token_accuracy": 6.385695887729525e-05,
      "num_tokens": 10855.0,
      "step": 5
    },
    {
      "entropy": 4.913383293151855,
      "epoch": 0.0009606147934678194,
      "grad_norm": 16.625,
      "learning_rate": 4.5e-06,
      "loss": 14.5216,
      "mean_token_accuracy": 0.00014638935390394182,
      "num_tokens": 24110.0,
      "step": 10
    },
    {
      "entropy": 4.925370550155639,
      "epoch": 0.001440922190201729,
      "grad_norm": 18.125,
      "learning_rate": 7e-06,
      "loss": 14.412,
      "mean_token_accuracy": 6.793478387407958e-05,
      "num_tokens": 35984.0,
      "step": 15
    },
    {
      "entropy": 4.947782182693482,
      "epoch": 0.0019212295869356388,
      "grad_norm": 19.5,
      "learning_rate": 9.5e-06,
      "loss": 14.2345,
      "mean_token_accuracy": 9.48766595683992e-05,
      "num_tokens": 48152.0,
      "step": 20
    },
    {
      "entropy": 5.064438819885254,
      "epoch": 0.0024015369836695487,
      "grad_norm": 23.75,
      "learning_rate": 1.2e-05,
      "loss": 13.7842,
      "mean_token_accuracy": 0.00013405063073150812,
      "num_tokens": 59810.0,
      "step": 25
    },
    {
      "entropy": 5.404191589355468,
      "epoch": 0.002881844380403458,
      "grad_norm": 31.125,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 13.1568,
      "mean_token_accuracy": 8.628127980045975e-05,
      "num_tokens": 70852.0,
      "step": 30
    },
    {
      "entropy": 7.142218971252442,
      "epoch": 0.0033621517771373678,
      "grad_norm": 21.0,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 11.9814,
      "mean_token_accuracy": 5.5741361575201156e-05,
      "num_tokens": 83378.0,
      "step": 35
    },
    {
      "entropy": 10.262308692932129,
      "epoch": 0.0038424591738712775,
      "grad_norm": 3.5625,
      "learning_rate": 1.95e-05,
      "loss": 10.9117,
      "mean_token_accuracy": 9.433962404727935e-05,
      "num_tokens": 95505.0,
      "step": 40
    },
    {
      "entropy": 10.698264122009277,
      "epoch": 0.004322766570605188,
      "grad_norm": 3.203125,
      "learning_rate": 2.2e-05,
      "loss": 10.7241,
      "mean_token_accuracy": 0.0005975825013592839,
      "num_tokens": 106812.0,
      "step": 45
    },
    {
      "entropy": 10.737651824951172,
      "epoch": 0.004803073967339097,
      "grad_norm": 2.78125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.5853,
      "mean_token_accuracy": 0.009875867154914886,
      "num_tokens": 118572.0,
      "step": 50
    },
    {
      "entropy": 10.73659906387329,
      "epoch": 0.005283381364073006,
      "grad_norm": 2.109375,
      "learning_rate": 2.7e-05,
      "loss": 10.3833,
      "mean_token_accuracy": 0.014529264252632856,
      "num_tokens": 130051.0,
      "step": 55
    },
    {
      "entropy": 10.701386642456054,
      "epoch": 0.005763688760806916,
      "grad_norm": 2.046875,
      "learning_rate": 2.95e-05,
      "loss": 10.2322,
      "mean_token_accuracy": 0.014093486312776803,
      "num_tokens": 141920.0,
      "step": 60
    },
    {
      "entropy": 10.711284351348876,
      "epoch": 0.006243996157540826,
      "grad_norm": 1.796875,
      "learning_rate": 3.2e-05,
      "loss": 10.045,
      "mean_token_accuracy": 0.015402647480368615,
      "num_tokens": 152706.0,
      "step": 65
    },
    {
      "entropy": 10.732075595855713,
      "epoch": 0.0067243035542747355,
      "grad_norm": 1.8125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.9817,
      "mean_token_accuracy": 0.013764715194702149,
      "num_tokens": 165253.0,
      "step": 70
    },
    {
      "entropy": 10.734086227416991,
      "epoch": 0.007204610951008645,
      "grad_norm": 1.8046875,
      "learning_rate": 3.7e-05,
      "loss": 9.8756,
      "mean_token_accuracy": 0.02153849443420768,
      "num_tokens": 176708.0,
      "step": 75
    },
    {
      "entropy": 10.729099750518799,
      "epoch": 0.007684918347742555,
      "grad_norm": 1.8125,
      "learning_rate": 3.95e-05,
      "loss": 9.8268,
      "mean_token_accuracy": 0.020801819767802952,
      "num_tokens": 188240.0,
      "step": 80
    },
    {
      "entropy": 10.723050022125244,
      "epoch": 0.008165225744476465,
      "grad_norm": 1.796875,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.6525,
      "mean_token_accuracy": 0.01858828719705343,
      "num_tokens": 198355.0,
      "step": 85
    },
    {
      "entropy": 10.719482040405273,
      "epoch": 0.008645533141210375,
      "grad_norm": 1.71875,
      "learning_rate": 4.45e-05,
      "loss": 9.5811,
      "mean_token_accuracy": 0.01679358258843422,
      "num_tokens": 209497.0,
      "step": 90
    },
    {
      "entropy": 10.711874198913574,
      "epoch": 0.009125840537944284,
      "grad_norm": 1.828125,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.5018,
      "mean_token_accuracy": 0.016100958082824947,
      "num_tokens": 220859.0,
      "step": 95
    },
    {
      "entropy": 10.695251655578613,
      "epoch": 0.009606147934678195,
      "grad_norm": 1.703125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.4214,
      "mean_token_accuracy": 0.017460816819220782,
      "num_tokens": 231550.0,
      "step": 100
    },
    {
      "entropy": 10.673291873931884,
      "epoch": 0.010086455331412104,
      "grad_norm": 1.671875,
      "learning_rate": 5.2e-05,
      "loss": 9.2854,
      "mean_token_accuracy": 0.016554870828986167,
      "num_tokens": 244210.0,
      "step": 105
    },
    {
      "entropy": 10.652278709411622,
      "epoch": 0.010566762728146013,
      "grad_norm": 1.7578125,
      "learning_rate": 5.45e-05,
      "loss": 9.1581,
      "mean_token_accuracy": 0.024293087422847748,
      "num_tokens": 255745.0,
      "step": 110
    },
    {
      "entropy": 10.617483043670655,
      "epoch": 0.011047070124879923,
      "grad_norm": 1.5703125,
      "learning_rate": 5.7e-05,
      "loss": 9.055,
      "mean_token_accuracy": 0.03160325270146132,
      "num_tokens": 266180.0,
      "step": 115
    },
    {
      "entropy": 10.569834327697754,
      "epoch": 0.011527377521613832,
      "grad_norm": 1.65625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.939,
      "mean_token_accuracy": 0.025451960042119026,
      "num_tokens": 277736.0,
      "step": 120
    },
    {
      "entropy": 10.52544527053833,
      "epoch": 0.012007684918347743,
      "grad_norm": 1.59375,
      "learning_rate": 6.2e-05,
      "loss": 8.8203,
      "mean_token_accuracy": 0.034478765353560446,
      "num_tokens": 289069.0,
      "step": 125
    },
    {
      "entropy": 10.468996620178222,
      "epoch": 0.012487992315081652,
      "grad_norm": 1.5546875,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.6972,
      "mean_token_accuracy": 0.03924299031496048,
      "num_tokens": 300240.0,
      "step": 130
    },
    {
      "entropy": 10.390470600128173,
      "epoch": 0.012968299711815562,
      "grad_norm": 1.5390625,
      "learning_rate": 6.7e-05,
      "loss": 8.622,
      "mean_token_accuracy": 0.033455712348222734,
      "num_tokens": 311698.0,
      "step": 135
    },
    {
      "entropy": 10.308708190917969,
      "epoch": 0.013448607108549471,
      "grad_norm": 1.359375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.4986,
      "mean_token_accuracy": 0.036465326137840745,
      "num_tokens": 322844.0,
      "step": 140
    },
    {
      "entropy": 10.209622859954834,
      "epoch": 0.013928914505283382,
      "grad_norm": 1.296875,
      "learning_rate": 7.2e-05,
      "loss": 8.3864,
      "mean_token_accuracy": 0.04093883950263262,
      "num_tokens": 335092.0,
      "step": 145
    },
    {
      "entropy": 10.091598796844483,
      "epoch": 0.01440922190201729,
      "grad_norm": 1.2578125,
      "learning_rate": 7.45e-05,
      "loss": 8.2513,
      "mean_token_accuracy": 0.041372421011328694,
      "num_tokens": 347033.0,
      "step": 150
    },
    {
      "entropy": 9.934578609466552,
      "epoch": 0.014889529298751201,
      "grad_norm": 1.203125,
      "learning_rate": 7.7e-05,
      "loss": 8.133,
      "mean_token_accuracy": 0.04138694703578949,
      "num_tokens": 358696.0,
      "step": 155
    },
    {
      "entropy": 9.762340259552001,
      "epoch": 0.01536983669548511,
      "grad_norm": 1.1328125,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.0344,
      "mean_token_accuracy": 0.04901535548269749,
      "num_tokens": 369390.0,
      "step": 160
    },
    {
      "entropy": 9.557822990417481,
      "epoch": 0.01585014409221902,
      "grad_norm": 0.984375,
      "learning_rate": 8.2e-05,
      "loss": 7.9344,
      "mean_token_accuracy": 0.04700087420642376,
      "num_tokens": 380540.0,
      "step": 165
    },
    {
      "entropy": 9.355019855499268,
      "epoch": 0.01633045148895293,
      "grad_norm": 1.015625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.8587,
      "mean_token_accuracy": 0.04431998413056135,
      "num_tokens": 391243.0,
      "step": 170
    },
    {
      "entropy": 9.114866542816163,
      "epoch": 0.01681075888568684,
      "grad_norm": 0.8671875,
      "learning_rate": 8.7e-05,
      "loss": 7.7947,
      "mean_token_accuracy": 0.05138894282281399,
      "num_tokens": 403336.0,
      "step": 175
    },
    {
      "entropy": 8.863803100585937,
      "epoch": 0.01729106628242075,
      "grad_norm": 0.83984375,
      "learning_rate": 8.95e-05,
      "loss": 7.693,
      "mean_token_accuracy": 0.054227447509765624,
      "num_tokens": 413886.0,
      "step": 180
    },
    {
      "entropy": 8.668179988861084,
      "epoch": 0.01777137367915466,
      "grad_norm": 0.95703125,
      "learning_rate": 9.2e-05,
      "loss": 7.7008,
      "mean_token_accuracy": 0.05373974815011025,
      "num_tokens": 425277.0,
      "step": 185
    },
    {
      "entropy": 8.488366603851318,
      "epoch": 0.01825168107588857,
      "grad_norm": 0.72265625,
      "learning_rate": 9.45e-05,
      "loss": 7.7708,
      "mean_token_accuracy": 0.05546994470059872,
      "num_tokens": 436868.0,
      "step": 190
    },
    {
      "entropy": 8.390744590759278,
      "epoch": 0.018731988472622477,
      "grad_norm": 1.078125,
      "learning_rate": 9.7e-05,
      "loss": 7.5877,
      "mean_token_accuracy": 0.052850323542952535,
      "num_tokens": 448349.0,
      "step": 195
    },
    {
      "entropy": 8.283522891998292,
      "epoch": 0.01921229586935639,
      "grad_norm": 0.77734375,
      "learning_rate": 9.95e-05,
      "loss": 7.5809,
      "mean_token_accuracy": 0.05953923240303993,
      "num_tokens": 459447.0,
      "step": 200
    },
    {
      "entropy": 8.150887298583985,
      "epoch": 0.0196926032660903,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000102,
      "loss": 7.5463,
      "mean_token_accuracy": 0.05646048374474048,
      "num_tokens": 470734.0,
      "step": 205
    },
    {
      "entropy": 8.137212657928467,
      "epoch": 0.020172910662824207,
      "grad_norm": 0.7578125,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.5847,
      "mean_token_accuracy": 0.06030619703233242,
      "num_tokens": 482015.0,
      "step": 210
    },
    {
      "entropy": 8.123589897155762,
      "epoch": 0.020653218059558116,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000107,
      "loss": 7.5311,
      "mean_token_accuracy": 0.06505471915006637,
      "num_tokens": 493339.0,
      "step": 215
    },
    {
      "entropy": 8.091688442230225,
      "epoch": 0.021133525456292025,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0001095,
      "loss": 7.4806,
      "mean_token_accuracy": 0.06740148589015008,
      "num_tokens": 504924.0,
      "step": 220
    },
    {
      "entropy": 8.105245876312257,
      "epoch": 0.021613832853025938,
      "grad_norm": 1.25,
      "learning_rate": 0.000112,
      "loss": 7.4493,
      "mean_token_accuracy": 0.06430868282914162,
      "num_tokens": 516603.0,
      "step": 225
    },
    {
      "entropy": 8.071302318572998,
      "epoch": 0.022094140249759846,
      "grad_norm": 1.1875,
      "learning_rate": 0.0001145,
      "loss": 7.3628,
      "mean_token_accuracy": 0.07100334689021111,
      "num_tokens": 528347.0,
      "step": 230
    },
    {
      "entropy": 8.018728876113892,
      "epoch": 0.022574447646493755,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.4365,
      "mean_token_accuracy": 0.07649173960089684,
      "num_tokens": 539328.0,
      "step": 235
    },
    {
      "entropy": 8.024458360671996,
      "epoch": 0.023054755043227664,
      "grad_norm": 0.921875,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.4121,
      "mean_token_accuracy": 0.07992132529616355,
      "num_tokens": 549297.0,
      "step": 240
    },
    {
      "entropy": 7.979285478591919,
      "epoch": 0.023535062439961577,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000122,
      "loss": 7.373,
      "mean_token_accuracy": 0.07391561307013035,
      "num_tokens": 560306.0,
      "step": 245
    },
    {
      "entropy": 8.05026822090149,
      "epoch": 0.024015369836695485,
      "grad_norm": 2.078125,
      "learning_rate": 0.0001245,
      "loss": 7.4298,
      "mean_token_accuracy": 0.07491427883505822,
      "num_tokens": 571972.0,
      "step": 250
    },
    {
      "entropy": 8.011105632781982,
      "epoch": 0.024495677233429394,
      "grad_norm": 0.984375,
      "learning_rate": 0.000127,
      "loss": 7.3324,
      "mean_token_accuracy": 0.07995120882987976,
      "num_tokens": 582962.0,
      "step": 255
    },
    {
      "entropy": 7.978642559051513,
      "epoch": 0.024975984630163303,
      "grad_norm": 1.140625,
      "learning_rate": 0.0001295,
      "loss": 7.29,
      "mean_token_accuracy": 0.07967569157481194,
      "num_tokens": 597193.0,
      "step": 260
    },
    {
      "entropy": 7.959802865982056,
      "epoch": 0.025456292026897216,
      "grad_norm": 1.5625,
      "learning_rate": 0.000132,
      "loss": 7.3629,
      "mean_token_accuracy": 0.08052113763988018,
      "num_tokens": 608982.0,
      "step": 265
    },
    {
      "entropy": 7.93332347869873,
      "epoch": 0.025936599423631124,
      "grad_norm": 1.265625,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.3853,
      "mean_token_accuracy": 0.07839832231402397,
      "num_tokens": 619953.0,
      "step": 270
    },
    {
      "entropy": 8.012384223937989,
      "epoch": 0.026416906820365033,
      "grad_norm": 1.46875,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.3273,
      "mean_token_accuracy": 0.08211004585027695,
      "num_tokens": 631039.0,
      "step": 275
    },
    {
      "entropy": 7.914879417419433,
      "epoch": 0.026897214217098942,
      "grad_norm": 1.375,
      "learning_rate": 0.0001395,
      "loss": 7.3281,
      "mean_token_accuracy": 0.0799741841852665,
      "num_tokens": 642656.0,
      "step": 280
    },
    {
      "entropy": 7.884822368621826,
      "epoch": 0.027377521613832854,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.1901,
      "mean_token_accuracy": 0.09594471827149391,
      "num_tokens": 653748.0,
      "step": 285
    },
    {
      "entropy": 7.860710859298706,
      "epoch": 0.027857829010566763,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0001445,
      "loss": 7.1763,
      "mean_token_accuracy": 0.08746851906180382,
      "num_tokens": 665618.0,
      "step": 290
    },
    {
      "entropy": 7.799470329284668,
      "epoch": 0.028338136407300672,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000147,
      "loss": 7.1972,
      "mean_token_accuracy": 0.08807335570454597,
      "num_tokens": 677329.0,
      "step": 295
    },
    {
      "entropy": 7.814197397232055,
      "epoch": 0.02881844380403458,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0001495,
      "loss": 7.1466,
      "mean_token_accuracy": 0.09343294948339462,
      "num_tokens": 688278.0,
      "step": 300
    },
    {
      "entropy": 7.766972303390503,
      "epoch": 0.029298751200768493,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000152,
      "loss": 7.2009,
      "mean_token_accuracy": 0.09057235419750213,
      "num_tokens": 700739.0,
      "step": 305
    },
    {
      "entropy": 7.782851505279541,
      "epoch": 0.029779058597502402,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.1242,
      "mean_token_accuracy": 0.09343240261077881,
      "num_tokens": 712527.0,
      "step": 310
    },
    {
      "entropy": 7.725392866134643,
      "epoch": 0.03025936599423631,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000157,
      "loss": 7.1564,
      "mean_token_accuracy": 0.09019288495182991,
      "num_tokens": 724514.0,
      "step": 315
    },
    {
      "entropy": 7.741910028457641,
      "epoch": 0.03073967339097022,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0001595,
      "loss": 7.1047,
      "mean_token_accuracy": 0.09378963932394982,
      "num_tokens": 735679.0,
      "step": 320
    },
    {
      "entropy": 7.70381178855896,
      "epoch": 0.03121998078770413,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000162,
      "loss": 7.0993,
      "mean_token_accuracy": 0.0981727458536625,
      "num_tokens": 747896.0,
      "step": 325
    },
    {
      "entropy": 7.6704418659210205,
      "epoch": 0.03170028818443804,
      "grad_norm": 1.40625,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.1443,
      "mean_token_accuracy": 0.09557413160800934,
      "num_tokens": 759081.0,
      "step": 330
    },
    {
      "entropy": 7.715982484817505,
      "epoch": 0.03218059558117195,
      "grad_norm": 1.546875,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.0815,
      "mean_token_accuracy": 0.09094736129045486,
      "num_tokens": 770459.0,
      "step": 335
    },
    {
      "entropy": 7.668689107894897,
      "epoch": 0.03266090297790586,
      "grad_norm": 1.078125,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.0395,
      "mean_token_accuracy": 0.09340968057513237,
      "num_tokens": 783960.0,
      "step": 340
    },
    {
      "entropy": 7.586823654174805,
      "epoch": 0.03314121037463977,
      "grad_norm": 1.34375,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.118,
      "mean_token_accuracy": 0.10006800666451454,
      "num_tokens": 795425.0,
      "step": 345
    },
    {
      "entropy": 7.6425714015960695,
      "epoch": 0.03362151777137368,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.0715,
      "mean_token_accuracy": 0.09306764900684357,
      "num_tokens": 807536.0,
      "step": 350
    },
    {
      "entropy": 7.6408501148223875,
      "epoch": 0.034101825168107586,
      "grad_norm": 1.171875,
      "learning_rate": 0.000177,
      "loss": 7.1133,
      "mean_token_accuracy": 0.09133664220571518,
      "num_tokens": 818801.0,
      "step": 355
    },
    {
      "entropy": 7.592654371261597,
      "epoch": 0.0345821325648415,
      "grad_norm": 1.171875,
      "learning_rate": 0.0001795,
      "loss": 6.997,
      "mean_token_accuracy": 0.09742675051093101,
      "num_tokens": 831497.0,
      "step": 360
    },
    {
      "entropy": 7.561050844192505,
      "epoch": 0.03506243996157541,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000182,
      "loss": 6.9518,
      "mean_token_accuracy": 0.10334301143884658,
      "num_tokens": 842491.0,
      "step": 365
    },
    {
      "entropy": 7.5975159168243405,
      "epoch": 0.03554274735830932,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0001845,
      "loss": 7.038,
      "mean_token_accuracy": 0.09360877349972725,
      "num_tokens": 854560.0,
      "step": 370
    },
    {
      "entropy": 7.535368490219116,
      "epoch": 0.03602305475504323,
      "grad_norm": 1.140625,
      "learning_rate": 0.000187,
      "loss": 6.9465,
      "mean_token_accuracy": 0.1063318483531475,
      "num_tokens": 866688.0,
      "step": 375
    },
    {
      "entropy": 7.5550884246826175,
      "epoch": 0.03650336215177714,
      "grad_norm": 1.515625,
      "learning_rate": 0.0001895,
      "loss": 7.0347,
      "mean_token_accuracy": 0.09658555164933205,
      "num_tokens": 879484.0,
      "step": 380
    },
    {
      "entropy": 7.572937870025635,
      "epoch": 0.036983669548511046,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000192,
      "loss": 7.0152,
      "mean_token_accuracy": 0.10198549181222916,
      "num_tokens": 890807.0,
      "step": 385
    },
    {
      "entropy": 7.576615810394287,
      "epoch": 0.037463976945244955,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0001945,
      "loss": 6.9605,
      "mean_token_accuracy": 0.10429177060723305,
      "num_tokens": 901759.0,
      "step": 390
    },
    {
      "entropy": 7.430867433547974,
      "epoch": 0.037944284341978864,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.89,
      "mean_token_accuracy": 0.10844507589936256,
      "num_tokens": 912212.0,
      "step": 395
    },
    {
      "entropy": 7.53248782157898,
      "epoch": 0.03842459173871278,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.9851,
      "mean_token_accuracy": 0.10315561667084694,
      "num_tokens": 923947.0,
      "step": 400
    },
    {
      "entropy": 7.401272916793824,
      "epoch": 0.03890489913544669,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000202,
      "loss": 6.9312,
      "mean_token_accuracy": 0.10761535987257957,
      "num_tokens": 935732.0,
      "step": 405
    },
    {
      "entropy": 7.374160099029541,
      "epoch": 0.0393852065321806,
      "grad_norm": 1.296875,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.8693,
      "mean_token_accuracy": 0.10408192500472069,
      "num_tokens": 948261.0,
      "step": 410
    },
    {
      "entropy": 7.473669195175171,
      "epoch": 0.039865513928914506,
      "grad_norm": 1.375,
      "learning_rate": 0.000207,
      "loss": 6.9775,
      "mean_token_accuracy": 0.10531274750828742,
      "num_tokens": 959574.0,
      "step": 415
    },
    {
      "entropy": 7.502210903167724,
      "epoch": 0.040345821325648415,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002095,
      "loss": 7.0212,
      "mean_token_accuracy": 0.10413776263594628,
      "num_tokens": 970329.0,
      "step": 420
    },
    {
      "entropy": 7.457794141769409,
      "epoch": 0.040826128722382324,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000212,
      "loss": 6.897,
      "mean_token_accuracy": 0.11288983076810837,
      "num_tokens": 982037.0,
      "step": 425
    },
    {
      "entropy": 7.320314025878906,
      "epoch": 0.04130643611911623,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002145,
      "loss": 6.8488,
      "mean_token_accuracy": 0.10890574008226395,
      "num_tokens": 994612.0,
      "step": 430
    },
    {
      "entropy": 7.436053037643433,
      "epoch": 0.04178674351585014,
      "grad_norm": 0.96875,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.9562,
      "mean_token_accuracy": 0.11320589855313301,
      "num_tokens": 1005960.0,
      "step": 435
    },
    {
      "entropy": 7.342877960205078,
      "epoch": 0.04226705091258405,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002195,
      "loss": 6.8312,
      "mean_token_accuracy": 0.10930031016469002,
      "num_tokens": 1017618.0,
      "step": 440
    },
    {
      "entropy": 7.3924178123474125,
      "epoch": 0.042747358309317966,
      "grad_norm": 1.78125,
      "learning_rate": 0.000222,
      "loss": 6.8722,
      "mean_token_accuracy": 0.10684446468949318,
      "num_tokens": 1029307.0,
      "step": 445
    },
    {
      "entropy": 7.34854736328125,
      "epoch": 0.043227665706051875,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002245,
      "loss": 6.8798,
      "mean_token_accuracy": 0.10711006596684455,
      "num_tokens": 1042027.0,
      "step": 450
    },
    {
      "entropy": 7.3988017559051515,
      "epoch": 0.043707973102785784,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.8815,
      "mean_token_accuracy": 0.10970013663172722,
      "num_tokens": 1053125.0,
      "step": 455
    },
    {
      "entropy": 7.270225954055786,
      "epoch": 0.04418828049951969,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.8643,
      "mean_token_accuracy": 0.11234270855784416,
      "num_tokens": 1064908.0,
      "step": 460
    },
    {
      "entropy": 7.329568862915039,
      "epoch": 0.0446685878962536,
      "grad_norm": 1.125,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.7486,
      "mean_token_accuracy": 0.11601889058947563,
      "num_tokens": 1076328.0,
      "step": 465
    },
    {
      "entropy": 7.339612579345703,
      "epoch": 0.04514889529298751,
      "grad_norm": 1.546875,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.9582,
      "mean_token_accuracy": 0.10401691868901253,
      "num_tokens": 1088469.0,
      "step": 470
    },
    {
      "entropy": 7.294072389602661,
      "epoch": 0.04562920268972142,
      "grad_norm": 1.46875,
      "learning_rate": 0.000237,
      "loss": 6.764,
      "mean_token_accuracy": 0.11554281786084175,
      "num_tokens": 1099408.0,
      "step": 475
    },
    {
      "entropy": 7.170223951339722,
      "epoch": 0.04610951008645533,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002395,
      "loss": 6.6937,
      "mean_token_accuracy": 0.11718685477972031,
      "num_tokens": 1111101.0,
      "step": 480
    },
    {
      "entropy": 7.192569828033447,
      "epoch": 0.046589817483189244,
      "grad_norm": 1.4375,
      "learning_rate": 0.000242,
      "loss": 6.6999,
      "mean_token_accuracy": 0.11837229505181313,
      "num_tokens": 1122877.0,
      "step": 485
    },
    {
      "entropy": 7.259336137771607,
      "epoch": 0.04707012487992315,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002445,
      "loss": 6.811,
      "mean_token_accuracy": 0.10224953144788743,
      "num_tokens": 1133956.0,
      "step": 490
    },
    {
      "entropy": 7.205342721939087,
      "epoch": 0.04755043227665706,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000247,
      "loss": 6.7171,
      "mean_token_accuracy": 0.11888269484043121,
      "num_tokens": 1146101.0,
      "step": 495
    },
    {
      "entropy": 7.218681526184082,
      "epoch": 0.04803073967339097,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002495,
      "loss": 6.8062,
      "mean_token_accuracy": 0.12109306752681732,
      "num_tokens": 1157432.0,
      "step": 500
    },
    {
      "entropy": 7.241644239425659,
      "epoch": 0.04851104707012488,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000252,
      "loss": 6.7906,
      "mean_token_accuracy": 0.11429283022880554,
      "num_tokens": 1167601.0,
      "step": 505
    },
    {
      "entropy": 7.20673680305481,
      "epoch": 0.04899135446685879,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002545,
      "loss": 6.6849,
      "mean_token_accuracy": 0.11796795651316642,
      "num_tokens": 1178818.0,
      "step": 510
    },
    {
      "entropy": 7.338191556930542,
      "epoch": 0.0494716618635927,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000257,
      "loss": 6.9209,
      "mean_token_accuracy": 0.11272176653146744,
      "num_tokens": 1189977.0,
      "step": 515
    },
    {
      "entropy": 7.15383129119873,
      "epoch": 0.049951969260326606,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002595,
      "loss": 6.7585,
      "mean_token_accuracy": 0.11906541213393211,
      "num_tokens": 1201039.0,
      "step": 520
    },
    {
      "entropy": 7.194189023971558,
      "epoch": 0.05043227665706052,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000262,
      "loss": 6.7788,
      "mean_token_accuracy": 0.11931035146117211,
      "num_tokens": 1212573.0,
      "step": 525
    },
    {
      "entropy": 7.264106559753418,
      "epoch": 0.05091258405379443,
      "grad_norm": 1.25,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.7181,
      "mean_token_accuracy": 0.12056513205170631,
      "num_tokens": 1223382.0,
      "step": 530
    },
    {
      "entropy": 7.170845222473145,
      "epoch": 0.05139289145052834,
      "grad_norm": 1.203125,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.7927,
      "mean_token_accuracy": 0.1244619682431221,
      "num_tokens": 1236501.0,
      "step": 535
    },
    {
      "entropy": 7.22953052520752,
      "epoch": 0.05187319884726225,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.7818,
      "mean_token_accuracy": 0.11887621656060218,
      "num_tokens": 1246798.0,
      "step": 540
    },
    {
      "entropy": 7.1261049747467045,
      "epoch": 0.05235350624399616,
      "grad_norm": 1.4375,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.7113,
      "mean_token_accuracy": 0.12007794082164765,
      "num_tokens": 1258182.0,
      "step": 545
    },
    {
      "entropy": 7.136359167098999,
      "epoch": 0.052833813640730067,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002745,
      "loss": 6.6647,
      "mean_token_accuracy": 0.11331228986382484,
      "num_tokens": 1270273.0,
      "step": 550
    },
    {
      "entropy": 7.1538010597229,
      "epoch": 0.053314121037463975,
      "grad_norm": 1.5625,
      "learning_rate": 0.000277,
      "loss": 6.7157,
      "mean_token_accuracy": 0.1148271806538105,
      "num_tokens": 1281136.0,
      "step": 555
    },
    {
      "entropy": 7.182579135894775,
      "epoch": 0.053794428434197884,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002795,
      "loss": 6.6991,
      "mean_token_accuracy": 0.11754038706421852,
      "num_tokens": 1294488.0,
      "step": 560
    },
    {
      "entropy": 7.084699916839599,
      "epoch": 0.05427473583093179,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.6054,
      "mean_token_accuracy": 0.12230072245001793,
      "num_tokens": 1304113.0,
      "step": 565
    },
    {
      "entropy": 7.104630041122436,
      "epoch": 0.05475504322766571,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002845,
      "loss": 6.6713,
      "mean_token_accuracy": 0.12584275975823403,
      "num_tokens": 1315417.0,
      "step": 570
    },
    {
      "entropy": 7.049445104598999,
      "epoch": 0.05523535062439962,
      "grad_norm": 1.359375,
      "learning_rate": 0.000287,
      "loss": 6.6306,
      "mean_token_accuracy": 0.12282092794775963,
      "num_tokens": 1328084.0,
      "step": 575
    },
    {
      "entropy": 7.128454923629761,
      "epoch": 0.05571565802113353,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002895,
      "loss": 6.7747,
      "mean_token_accuracy": 0.12278011664748192,
      "num_tokens": 1338696.0,
      "step": 580
    },
    {
      "entropy": 7.090286588668823,
      "epoch": 0.056195965417867436,
      "grad_norm": 1.078125,
      "learning_rate": 0.000292,
      "loss": 6.7179,
      "mean_token_accuracy": 0.11662147119641304,
      "num_tokens": 1350240.0,
      "step": 585
    },
    {
      "entropy": 7.117737483978272,
      "epoch": 0.056676272814601344,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002945,
      "loss": 6.6639,
      "mean_token_accuracy": 0.11824186667799949,
      "num_tokens": 1361720.0,
      "step": 590
    },
    {
      "entropy": 6.979755496978759,
      "epoch": 0.05715658021133525,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000297,
      "loss": 6.6611,
      "mean_token_accuracy": 0.12714812606573106,
      "num_tokens": 1373286.0,
      "step": 595
    },
    {
      "entropy": 7.1012420654296875,
      "epoch": 0.05763688760806916,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002995,
      "loss": 6.6116,
      "mean_token_accuracy": 0.12019829899072647,
      "num_tokens": 1384274.0,
      "step": 600
    },
    {
      "entropy": 7.009186458587647,
      "epoch": 0.05811719500480307,
      "grad_norm": 1.484375,
      "learning_rate": 0.000302,
      "loss": 6.6625,
      "mean_token_accuracy": 0.12345866709947587,
      "num_tokens": 1395355.0,
      "step": 605
    },
    {
      "entropy": 6.984350728988647,
      "epoch": 0.05859750240153699,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003045,
      "loss": 6.5861,
      "mean_token_accuracy": 0.1286102719604969,
      "num_tokens": 1406664.0,
      "step": 610
    },
    {
      "entropy": 7.138849353790283,
      "epoch": 0.059077809798270896,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000307,
      "loss": 6.7404,
      "mean_token_accuracy": 0.11916537806391717,
      "num_tokens": 1418450.0,
      "step": 615
    },
    {
      "entropy": 7.0278784275054935,
      "epoch": 0.059558117195004805,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003095,
      "loss": 6.6374,
      "mean_token_accuracy": 0.1211201086640358,
      "num_tokens": 1430048.0,
      "step": 620
    },
    {
      "entropy": 7.098714780807495,
      "epoch": 0.060038424591738714,
      "grad_norm": 1.484375,
      "learning_rate": 0.000312,
      "loss": 6.6739,
      "mean_token_accuracy": 0.12407828345894814,
      "num_tokens": 1441820.0,
      "step": 625
    },
    {
      "entropy": 6.935336065292359,
      "epoch": 0.06051873198847262,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003145,
      "loss": 6.5317,
      "mean_token_accuracy": 0.12386323809623719,
      "num_tokens": 1453209.0,
      "step": 630
    },
    {
      "entropy": 6.9461079120635985,
      "epoch": 0.06099903938520653,
      "grad_norm": 1.046875,
      "learning_rate": 0.000317,
      "loss": 6.546,
      "mean_token_accuracy": 0.122261643409729,
      "num_tokens": 1465423.0,
      "step": 635
    },
    {
      "entropy": 6.992834091186523,
      "epoch": 0.06147934678194044,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003195,
      "loss": 6.5324,
      "mean_token_accuracy": 0.12468917295336723,
      "num_tokens": 1476575.0,
      "step": 640
    },
    {
      "entropy": 6.9027880191802975,
      "epoch": 0.06195965417867435,
      "grad_norm": 1.5,
      "learning_rate": 0.000322,
      "loss": 6.495,
      "mean_token_accuracy": 0.12275839298963546,
      "num_tokens": 1486932.0,
      "step": 645
    },
    {
      "entropy": 6.965601539611816,
      "epoch": 0.06243996157540826,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.532,
      "mean_token_accuracy": 0.12491905614733696,
      "num_tokens": 1498494.0,
      "step": 650
    },
    {
      "entropy": 6.929765415191651,
      "epoch": 0.06292026897214217,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.5786,
      "mean_token_accuracy": 0.1209442898631096,
      "num_tokens": 1509937.0,
      "step": 655
    },
    {
      "entropy": 7.01481876373291,
      "epoch": 0.06340057636887608,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.6063,
      "mean_token_accuracy": 0.1320799134671688,
      "num_tokens": 1519823.0,
      "step": 660
    },
    {
      "entropy": 7.075559997558594,
      "epoch": 0.06388088376560999,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.6301,
      "mean_token_accuracy": 0.12214003503322601,
      "num_tokens": 1529943.0,
      "step": 665
    },
    {
      "entropy": 6.827705144882202,
      "epoch": 0.0643611911623439,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.4589,
      "mean_token_accuracy": 0.12923690006136895,
      "num_tokens": 1540618.0,
      "step": 670
    },
    {
      "entropy": 6.919480466842652,
      "epoch": 0.06484149855907781,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000337,
      "loss": 6.5779,
      "mean_token_accuracy": 0.12186570391058922,
      "num_tokens": 1553208.0,
      "step": 675
    },
    {
      "entropy": 6.962458515167237,
      "epoch": 0.06532180595581172,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003395,
      "loss": 6.5472,
      "mean_token_accuracy": 0.13120265454053878,
      "num_tokens": 1563975.0,
      "step": 680
    },
    {
      "entropy": 7.013445615768433,
      "epoch": 0.06580211335254563,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000342,
      "loss": 6.5969,
      "mean_token_accuracy": 0.12027480229735374,
      "num_tokens": 1575998.0,
      "step": 685
    },
    {
      "entropy": 6.872586297988891,
      "epoch": 0.06628242074927954,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.4957,
      "mean_token_accuracy": 0.12978472113609313,
      "num_tokens": 1586041.0,
      "step": 690
    },
    {
      "entropy": 6.915009069442749,
      "epoch": 0.06676272814601344,
      "grad_norm": 1.546875,
      "learning_rate": 0.000347,
      "loss": 6.5464,
      "mean_token_accuracy": 0.12723507061600686,
      "num_tokens": 1597531.0,
      "step": 695
    },
    {
      "entropy": 6.859202480316162,
      "epoch": 0.06724303554274735,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003495,
      "loss": 6.5324,
      "mean_token_accuracy": 0.1253533720970154,
      "num_tokens": 1609255.0,
      "step": 700
    },
    {
      "entropy": 6.813373184204101,
      "epoch": 0.06772334293948126,
      "grad_norm": 1.34375,
      "learning_rate": 0.000352,
      "loss": 6.5279,
      "mean_token_accuracy": 0.1321180894970894,
      "num_tokens": 1621098.0,
      "step": 705
    },
    {
      "entropy": 6.89396710395813,
      "epoch": 0.06820365033621517,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003545,
      "loss": 6.4187,
      "mean_token_accuracy": 0.14093777388334275,
      "num_tokens": 1631941.0,
      "step": 710
    },
    {
      "entropy": 6.811427450180053,
      "epoch": 0.0686839577329491,
      "grad_norm": 1.15625,
      "learning_rate": 0.000357,
      "loss": 6.4234,
      "mean_token_accuracy": 0.13498233929276465,
      "num_tokens": 1643117.0,
      "step": 715
    },
    {
      "entropy": 6.800242614746094,
      "epoch": 0.069164265129683,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003595,
      "loss": 6.4627,
      "mean_token_accuracy": 0.13018925338983536,
      "num_tokens": 1653595.0,
      "step": 720
    },
    {
      "entropy": 6.83650918006897,
      "epoch": 0.06964457252641691,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000362,
      "loss": 6.5025,
      "mean_token_accuracy": 0.13135557770729064,
      "num_tokens": 1664495.0,
      "step": 725
    },
    {
      "entropy": 6.802758979797363,
      "epoch": 0.07012487992315082,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003645,
      "loss": 6.4803,
      "mean_token_accuracy": 0.1322146400809288,
      "num_tokens": 1674923.0,
      "step": 730
    },
    {
      "entropy": 6.83389892578125,
      "epoch": 0.07060518731988473,
      "grad_norm": 1.265625,
      "learning_rate": 0.000367,
      "loss": 6.3745,
      "mean_token_accuracy": 0.1347797356545925,
      "num_tokens": 1685904.0,
      "step": 735
    },
    {
      "entropy": 6.834182548522949,
      "epoch": 0.07108549471661864,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003695,
      "loss": 6.5992,
      "mean_token_accuracy": 0.12240826413035392,
      "num_tokens": 1699133.0,
      "step": 740
    },
    {
      "entropy": 6.7865643978118895,
      "epoch": 0.07156580211335255,
      "grad_norm": 1.046875,
      "learning_rate": 0.000372,
      "loss": 6.408,
      "mean_token_accuracy": 0.13333356380462646,
      "num_tokens": 1711559.0,
      "step": 745
    },
    {
      "entropy": 6.8253981590271,
      "epoch": 0.07204610951008646,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003745,
      "loss": 6.3276,
      "mean_token_accuracy": 0.1389731504023075,
      "num_tokens": 1722526.0,
      "step": 750
    },
    {
      "entropy": 6.7026113986969,
      "epoch": 0.07252641690682037,
      "grad_norm": 1.328125,
      "learning_rate": 0.000377,
      "loss": 6.3874,
      "mean_token_accuracy": 0.13594599291682244,
      "num_tokens": 1734261.0,
      "step": 755
    },
    {
      "entropy": 6.764468193054199,
      "epoch": 0.07300672430355427,
      "grad_norm": 1.375,
      "learning_rate": 0.0003795,
      "loss": 6.3728,
      "mean_token_accuracy": 0.14337323382496833,
      "num_tokens": 1745151.0,
      "step": 760
    },
    {
      "entropy": 6.874772214889527,
      "epoch": 0.07348703170028818,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000382,
      "loss": 6.5351,
      "mean_token_accuracy": 0.132026257365942,
      "num_tokens": 1755463.0,
      "step": 765
    },
    {
      "entropy": 6.830476856231689,
      "epoch": 0.07396733909702209,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003845,
      "loss": 6.4887,
      "mean_token_accuracy": 0.13159892857074737,
      "num_tokens": 1767717.0,
      "step": 770
    },
    {
      "entropy": 6.660166883468628,
      "epoch": 0.074447646493756,
      "grad_norm": 1.421875,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.321,
      "mean_token_accuracy": 0.1407831773161888,
      "num_tokens": 1779115.0,
      "step": 775
    },
    {
      "entropy": 6.616621780395508,
      "epoch": 0.07492795389048991,
      "grad_norm": 1.375,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.2814,
      "mean_token_accuracy": 0.13830310627818107,
      "num_tokens": 1789644.0,
      "step": 780
    },
    {
      "entropy": 6.793073272705078,
      "epoch": 0.07540826128722382,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.4934,
      "mean_token_accuracy": 0.13178734481334686,
      "num_tokens": 1800606.0,
      "step": 785
    },
    {
      "entropy": 6.812487077713013,
      "epoch": 0.07588856868395773,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.4962,
      "mean_token_accuracy": 0.130271115899086,
      "num_tokens": 1812168.0,
      "step": 790
    },
    {
      "entropy": 6.774988603591919,
      "epoch": 0.07636887608069164,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.4425,
      "mean_token_accuracy": 0.12558576837182045,
      "num_tokens": 1823830.0,
      "step": 795
    },
    {
      "entropy": 6.7873927593231205,
      "epoch": 0.07684918347742556,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003995,
      "loss": 6.4269,
      "mean_token_accuracy": 0.13118278980255127,
      "num_tokens": 1835611.0,
      "step": 800
    },
    {
      "entropy": 6.705799150466919,
      "epoch": 0.07732949087415947,
      "grad_norm": 1.328125,
      "learning_rate": 0.000402,
      "loss": 6.4715,
      "mean_token_accuracy": 0.12893390357494355,
      "num_tokens": 1847036.0,
      "step": 805
    },
    {
      "entropy": 6.743785047531128,
      "epoch": 0.07780979827089338,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004045,
      "loss": 6.416,
      "mean_token_accuracy": 0.1317111983895302,
      "num_tokens": 1857476.0,
      "step": 810
    },
    {
      "entropy": 6.710385322570801,
      "epoch": 0.07829010566762729,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.4346,
      "mean_token_accuracy": 0.13412694856524468,
      "num_tokens": 1869073.0,
      "step": 815
    },
    {
      "entropy": 6.6248006343841555,
      "epoch": 0.0787704130643612,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004095,
      "loss": 6.3166,
      "mean_token_accuracy": 0.1385165847837925,
      "num_tokens": 1880439.0,
      "step": 820
    },
    {
      "entropy": 6.801689147949219,
      "epoch": 0.0792507204610951,
      "grad_norm": 1.296875,
      "learning_rate": 0.000412,
      "loss": 6.499,
      "mean_token_accuracy": 0.128297408670187,
      "num_tokens": 1891600.0,
      "step": 825
    },
    {
      "entropy": 6.755054044723511,
      "epoch": 0.07973102785782901,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004145,
      "loss": 6.4488,
      "mean_token_accuracy": 0.1346999153494835,
      "num_tokens": 1903126.0,
      "step": 830
    },
    {
      "entropy": 6.787124109268189,
      "epoch": 0.08021133525456292,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000417,
      "loss": 6.417,
      "mean_token_accuracy": 0.1378346562385559,
      "num_tokens": 1913913.0,
      "step": 835
    },
    {
      "entropy": 6.613849306106568,
      "epoch": 0.08069164265129683,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004195,
      "loss": 6.3337,
      "mean_token_accuracy": 0.14059634506702423,
      "num_tokens": 1924961.0,
      "step": 840
    },
    {
      "entropy": 6.824811744689941,
      "epoch": 0.08117195004803074,
      "grad_norm": 1.421875,
      "learning_rate": 0.000422,
      "loss": 6.4201,
      "mean_token_accuracy": 0.12982920035719872,
      "num_tokens": 1936773.0,
      "step": 845
    },
    {
      "entropy": 6.610549211502075,
      "epoch": 0.08165225744476465,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004245,
      "loss": 6.3343,
      "mean_token_accuracy": 0.13738347217440605,
      "num_tokens": 1948190.0,
      "step": 850
    },
    {
      "entropy": 6.6286153316497805,
      "epoch": 0.08213256484149856,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000427,
      "loss": 6.3262,
      "mean_token_accuracy": 0.138392174243927,
      "num_tokens": 1960038.0,
      "step": 855
    },
    {
      "entropy": 6.762907218933106,
      "epoch": 0.08261287223823247,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004295,
      "loss": 6.4021,
      "mean_token_accuracy": 0.13606049939990045,
      "num_tokens": 1970535.0,
      "step": 860
    },
    {
      "entropy": 6.655164575576782,
      "epoch": 0.08309317963496637,
      "grad_norm": 1.59375,
      "learning_rate": 0.000432,
      "loss": 6.3764,
      "mean_token_accuracy": 0.13180122673511505,
      "num_tokens": 1981386.0,
      "step": 865
    },
    {
      "entropy": 6.615013980865479,
      "epoch": 0.08357348703170028,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004345,
      "loss": 6.2949,
      "mean_token_accuracy": 0.1441991925239563,
      "num_tokens": 1993196.0,
      "step": 870
    },
    {
      "entropy": 6.65190896987915,
      "epoch": 0.08405379442843419,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000437,
      "loss": 6.3309,
      "mean_token_accuracy": 0.14546860083937646,
      "num_tokens": 2004756.0,
      "step": 875
    },
    {
      "entropy": 6.595526599884034,
      "epoch": 0.0845341018251681,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004395,
      "loss": 6.2971,
      "mean_token_accuracy": 0.1399446427822113,
      "num_tokens": 2016020.0,
      "step": 880
    },
    {
      "entropy": 6.638343858718872,
      "epoch": 0.08501440922190202,
      "grad_norm": 1.203125,
      "learning_rate": 0.000442,
      "loss": 6.3364,
      "mean_token_accuracy": 0.14163062497973442,
      "num_tokens": 2027747.0,
      "step": 885
    },
    {
      "entropy": 6.649935054779053,
      "epoch": 0.08549471661863593,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004445,
      "loss": 6.3262,
      "mean_token_accuracy": 0.13975787982344628,
      "num_tokens": 2038841.0,
      "step": 890
    },
    {
      "entropy": 6.640672445297241,
      "epoch": 0.08597502401536984,
      "grad_norm": 1.21875,
      "learning_rate": 0.000447,
      "loss": 6.3547,
      "mean_token_accuracy": 0.13744548857212066,
      "num_tokens": 2049905.0,
      "step": 895
    },
    {
      "entropy": 6.644440841674805,
      "epoch": 0.08645533141210375,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.4266,
      "mean_token_accuracy": 0.13376614451408386,
      "num_tokens": 2062492.0,
      "step": 900
    },
    {
      "entropy": 6.674169445037842,
      "epoch": 0.08693563880883766,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.3634,
      "mean_token_accuracy": 0.13451617136597632,
      "num_tokens": 2073840.0,
      "step": 905
    },
    {
      "entropy": 6.615630722045898,
      "epoch": 0.08741594620557157,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.2784,
      "mean_token_accuracy": 0.13790287226438522,
      "num_tokens": 2085720.0,
      "step": 910
    },
    {
      "entropy": 6.676623725891114,
      "epoch": 0.08789625360230548,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.3559,
      "mean_token_accuracy": 0.14061954542994498,
      "num_tokens": 2096649.0,
      "step": 915
    },
    {
      "entropy": 6.63926134109497,
      "epoch": 0.08837656099903939,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.3586,
      "mean_token_accuracy": 0.1393554113805294,
      "num_tokens": 2109030.0,
      "step": 920
    },
    {
      "entropy": 6.6048328399658205,
      "epoch": 0.0888568683957733,
      "grad_norm": 1.40625,
      "learning_rate": 0.000462,
      "loss": 6.2757,
      "mean_token_accuracy": 0.142194165289402,
      "num_tokens": 2121384.0,
      "step": 925
    },
    {
      "entropy": 6.563203191757202,
      "epoch": 0.0893371757925072,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004645,
      "loss": 6.2225,
      "mean_token_accuracy": 0.14719335213303567,
      "num_tokens": 2131377.0,
      "step": 930
    },
    {
      "entropy": 6.577233839035034,
      "epoch": 0.08981748318924111,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000467,
      "loss": 6.4123,
      "mean_token_accuracy": 0.13454320505261422,
      "num_tokens": 2142364.0,
      "step": 935
    },
    {
      "entropy": 6.619031524658203,
      "epoch": 0.09029779058597502,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004695,
      "loss": 6.2849,
      "mean_token_accuracy": 0.13945741057395936,
      "num_tokens": 2153040.0,
      "step": 940
    },
    {
      "entropy": 6.542799091339111,
      "epoch": 0.09077809798270893,
      "grad_norm": 1.3125,
      "learning_rate": 0.000472,
      "loss": 6.3295,
      "mean_token_accuracy": 0.14165178686380386,
      "num_tokens": 2165571.0,
      "step": 945
    },
    {
      "entropy": 6.620811080932617,
      "epoch": 0.09125840537944284,
      "grad_norm": 1.375,
      "learning_rate": 0.0004745,
      "loss": 6.3003,
      "mean_token_accuracy": 0.13918614387512207,
      "num_tokens": 2177241.0,
      "step": 950
    },
    {
      "entropy": 6.539478206634522,
      "epoch": 0.09173871277617675,
      "grad_norm": 1.25,
      "learning_rate": 0.000477,
      "loss": 6.3211,
      "mean_token_accuracy": 0.135619805008173,
      "num_tokens": 2187475.0,
      "step": 955
    },
    {
      "entropy": 6.6213781356811525,
      "epoch": 0.09221902017291066,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004795,
      "loss": 6.3492,
      "mean_token_accuracy": 0.1400366671383381,
      "num_tokens": 2198185.0,
      "step": 960
    },
    {
      "entropy": 6.5712048530578615,
      "epoch": 0.09269932756964457,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000482,
      "loss": 6.3041,
      "mean_token_accuracy": 0.1407790169119835,
      "num_tokens": 2210404.0,
      "step": 965
    },
    {
      "entropy": 6.612986421585083,
      "epoch": 0.09317963496637849,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004845,
      "loss": 6.2737,
      "mean_token_accuracy": 0.13978574126958848,
      "num_tokens": 2222188.0,
      "step": 970
    },
    {
      "entropy": 6.56772894859314,
      "epoch": 0.0936599423631124,
      "grad_norm": 1.390625,
      "learning_rate": 0.000487,
      "loss": 6.333,
      "mean_token_accuracy": 0.13907585740089418,
      "num_tokens": 2233418.0,
      "step": 975
    },
    {
      "entropy": 6.512521982192993,
      "epoch": 0.0941402497598463,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004895,
      "loss": 6.2849,
      "mean_token_accuracy": 0.1478159360587597,
      "num_tokens": 2245053.0,
      "step": 980
    },
    {
      "entropy": 6.605170869827271,
      "epoch": 0.09462055715658022,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000492,
      "loss": 6.3158,
      "mean_token_accuracy": 0.14485108256340026,
      "num_tokens": 2256375.0,
      "step": 985
    },
    {
      "entropy": 6.405674791336059,
      "epoch": 0.09510086455331412,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004945,
      "loss": 6.1641,
      "mean_token_accuracy": 0.1475372150540352,
      "num_tokens": 2267074.0,
      "step": 990
    },
    {
      "entropy": 6.520063352584839,
      "epoch": 0.09558117195004803,
      "grad_norm": 1.234375,
      "learning_rate": 0.000497,
      "loss": 6.1987,
      "mean_token_accuracy": 0.1514969453215599,
      "num_tokens": 2277168.0,
      "step": 995
    },
    {
      "entropy": 6.435000085830689,
      "epoch": 0.09606147934678194,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995,
      "loss": 6.1878,
      "mean_token_accuracy": 0.15069199353456497,
      "num_tokens": 2288178.0,
      "step": 1000
    },
    {
      "entropy": 6.503354835510254,
      "epoch": 0.09654178674351585,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999999983283737,
      "loss": 6.1929,
      "mean_token_accuracy": 0.14548032358288765,
      "num_tokens": 2299765.0,
      "step": 1005
    },
    {
      "entropy": 6.603171157836914,
      "epoch": 0.09702209414024976,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004999999915373924,
      "loss": 6.3207,
      "mean_token_accuracy": 0.1360747143626213,
      "num_tokens": 2312047.0,
      "step": 1010
    },
    {
      "entropy": 6.503301382064819,
      "epoch": 0.09750240153698367,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004999999795225793,
      "loss": 6.2995,
      "mean_token_accuracy": 0.1352042004466057,
      "num_tokens": 2324118.0,
      "step": 1015
    },
    {
      "entropy": 6.585948276519775,
      "epoch": 0.09798270893371758,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004999999622839347,
      "loss": 6.2985,
      "mean_token_accuracy": 0.14105767011642456,
      "num_tokens": 2335171.0,
      "step": 1020
    },
    {
      "entropy": 6.513099575042725,
      "epoch": 0.09846301633045149,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004999999398214593,
      "loss": 6.1997,
      "mean_token_accuracy": 0.14366835802793504,
      "num_tokens": 2346338.0,
      "step": 1025
    },
    {
      "entropy": 6.442260217666626,
      "epoch": 0.0989433237271854,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004999999121351532,
      "loss": 6.2248,
      "mean_token_accuracy": 0.15347937047481536,
      "num_tokens": 2357185.0,
      "step": 1030
    },
    {
      "entropy": 6.52047381401062,
      "epoch": 0.0994236311239193,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999998792250173,
      "loss": 6.1793,
      "mean_token_accuracy": 0.15213955715298652,
      "num_tokens": 2368494.0,
      "step": 1035
    },
    {
      "entropy": 6.535188245773315,
      "epoch": 0.09990393852065321,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004999998410910524,
      "loss": 6.3719,
      "mean_token_accuracy": 0.13562044501304626,
      "num_tokens": 2380800.0,
      "step": 1040
    },
    {
      "entropy": 6.575578689575195,
      "epoch": 0.10038424591738712,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999997977332592,
      "loss": 6.3159,
      "mean_token_accuracy": 0.14009395092725754,
      "num_tokens": 2391753.0,
      "step": 1045
    },
    {
      "entropy": 6.376717662811279,
      "epoch": 0.10086455331412104,
      "grad_norm": 1.25,
      "learning_rate": 0.0004999997491516389,
      "loss": 6.1859,
      "mean_token_accuracy": 0.14269104301929475,
      "num_tokens": 2403324.0,
      "step": 1050
    },
    {
      "entropy": 6.512421941757202,
      "epoch": 0.10134486071085495,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999996953461925,
      "loss": 6.2805,
      "mean_token_accuracy": 0.1382668823003769,
      "num_tokens": 2414873.0,
      "step": 1055
    },
    {
      "entropy": 6.387165594100952,
      "epoch": 0.10182516810758886,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999996363169212,
      "loss": 6.0851,
      "mean_token_accuracy": 0.15157609730958937,
      "num_tokens": 2425308.0,
      "step": 1060
    },
    {
      "entropy": 6.382681798934937,
      "epoch": 0.10230547550432277,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999995720638266,
      "loss": 6.1202,
      "mean_token_accuracy": 0.15165834426879882,
      "num_tokens": 2436835.0,
      "step": 1065
    },
    {
      "entropy": 6.456397867202758,
      "epoch": 0.10278578290105668,
      "grad_norm": 1.21875,
      "learning_rate": 0.00049999950258691,
      "loss": 6.2256,
      "mean_token_accuracy": 0.14640265703201294,
      "num_tokens": 2446798.0,
      "step": 1070
    },
    {
      "entropy": 6.332650756835937,
      "epoch": 0.10326609029779059,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999994278861731,
      "loss": 6.1221,
      "mean_token_accuracy": 0.14805126786231995,
      "num_tokens": 2457308.0,
      "step": 1075
    },
    {
      "entropy": 6.4620646953582765,
      "epoch": 0.1037463976945245,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999993479616175,
      "loss": 6.1985,
      "mean_token_accuracy": 0.13535723462700844,
      "num_tokens": 2468917.0,
      "step": 1080
    },
    {
      "entropy": 6.465546655654907,
      "epoch": 0.1042267050912584,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004999992628132451,
      "loss": 6.2148,
      "mean_token_accuracy": 0.1429847039282322,
      "num_tokens": 2481363.0,
      "step": 1085
    },
    {
      "entropy": 6.456140518188477,
      "epoch": 0.10470701248799232,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999991724410582,
      "loss": 6.2043,
      "mean_token_accuracy": 0.14445993900299073,
      "num_tokens": 2493082.0,
      "step": 1090
    },
    {
      "entropy": 6.3651855945587155,
      "epoch": 0.10518731988472622,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999990768450583,
      "loss": 6.1666,
      "mean_token_accuracy": 0.14558956474065782,
      "num_tokens": 2503849.0,
      "step": 1095
    },
    {
      "entropy": 6.470709943771363,
      "epoch": 0.10566762728146013,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004999989760252482,
      "loss": 6.2132,
      "mean_token_accuracy": 0.1486193262040615,
      "num_tokens": 2514528.0,
      "step": 1100
    },
    {
      "entropy": 6.319680213928223,
      "epoch": 0.10614793467819404,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004999988699816299,
      "loss": 6.1932,
      "mean_token_accuracy": 0.1492076650261879,
      "num_tokens": 2524971.0,
      "step": 1105
    },
    {
      "entropy": 6.440478706359864,
      "epoch": 0.10662824207492795,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999987587142058,
      "loss": 6.0986,
      "mean_token_accuracy": 0.14526561200618743,
      "num_tokens": 2535674.0,
      "step": 1110
    },
    {
      "entropy": 6.500247812271118,
      "epoch": 0.10710854947166186,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999986422229789,
      "loss": 6.3347,
      "mean_token_accuracy": 0.13726388663053513,
      "num_tokens": 2547108.0,
      "step": 1115
    },
    {
      "entropy": 6.384337902069092,
      "epoch": 0.10758885686839577,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999985205079514,
      "loss": 6.1486,
      "mean_token_accuracy": 0.14365847408771515,
      "num_tokens": 2559474.0,
      "step": 1120
    },
    {
      "entropy": 6.2849139213562015,
      "epoch": 0.10806916426512968,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999983935691265,
      "loss": 5.9875,
      "mean_token_accuracy": 0.16158309876918792,
      "num_tokens": 2571264.0,
      "step": 1125
    },
    {
      "entropy": 6.373571968078613,
      "epoch": 0.10854947166186359,
      "grad_norm": 1.125,
      "learning_rate": 0.000499998261406507,
      "loss": 6.1607,
      "mean_token_accuracy": 0.14907533824443817,
      "num_tokens": 2583731.0,
      "step": 1130
    },
    {
      "entropy": 6.570598459243774,
      "epoch": 0.10902977905859751,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999981240200958,
      "loss": 6.209,
      "mean_token_accuracy": 0.14517664313316345,
      "num_tokens": 2595497.0,
      "step": 1135
    },
    {
      "entropy": 6.279880571365356,
      "epoch": 0.10951008645533142,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999979814098966,
      "loss": 6.1441,
      "mean_token_accuracy": 0.14655442014336587,
      "num_tokens": 2607358.0,
      "step": 1140
    },
    {
      "entropy": 6.362181329727173,
      "epoch": 0.10999039385206533,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999978335759121,
      "loss": 6.087,
      "mean_token_accuracy": 0.15378076806664467,
      "num_tokens": 2618936.0,
      "step": 1145
    },
    {
      "entropy": 6.368446207046508,
      "epoch": 0.11047070124879924,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999976805181461,
      "loss": 6.2243,
      "mean_token_accuracy": 0.14323752224445344,
      "num_tokens": 2631840.0,
      "step": 1150
    },
    {
      "entropy": 6.41338152885437,
      "epoch": 0.11095100864553314,
      "grad_norm": 1.25,
      "learning_rate": 0.000499997522236602,
      "loss": 6.1738,
      "mean_token_accuracy": 0.1520377591252327,
      "num_tokens": 2642412.0,
      "step": 1155
    },
    {
      "entropy": 6.376598930358886,
      "epoch": 0.11143131604226705,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999973587312837,
      "loss": 6.1544,
      "mean_token_accuracy": 0.14735933318734168,
      "num_tokens": 2653890.0,
      "step": 1160
    },
    {
      "entropy": 6.446410703659057,
      "epoch": 0.11191162343900096,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999971900021947,
      "loss": 6.2058,
      "mean_token_accuracy": 0.15076753124594688,
      "num_tokens": 2664888.0,
      "step": 1165
    },
    {
      "entropy": 6.336441469192505,
      "epoch": 0.11239193083573487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999970160493391,
      "loss": 6.1032,
      "mean_token_accuracy": 0.14862595051527022,
      "num_tokens": 2675550.0,
      "step": 1170
    },
    {
      "entropy": 6.2902812480926515,
      "epoch": 0.11287223823246878,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999968368727209,
      "loss": 6.1139,
      "mean_token_accuracy": 0.15255775302648544,
      "num_tokens": 2688022.0,
      "step": 1175
    },
    {
      "entropy": 6.3844489574432375,
      "epoch": 0.11335254562920269,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999966524723442,
      "loss": 6.116,
      "mean_token_accuracy": 0.1474123939871788,
      "num_tokens": 2698737.0,
      "step": 1180
    },
    {
      "entropy": 6.314852046966553,
      "epoch": 0.1138328530259366,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999964628482135,
      "loss": 6.0924,
      "mean_token_accuracy": 0.15541501641273497,
      "num_tokens": 2709844.0,
      "step": 1185
    },
    {
      "entropy": 6.342870473861694,
      "epoch": 0.1143131604226705,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999962680003328,
      "loss": 6.1408,
      "mean_token_accuracy": 0.1548864386975765,
      "num_tokens": 2720273.0,
      "step": 1190
    },
    {
      "entropy": 6.401251649856567,
      "epoch": 0.11479346781940442,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499996067928707,
      "loss": 6.1552,
      "mean_token_accuracy": 0.15096053779125213,
      "num_tokens": 2731354.0,
      "step": 1195
    },
    {
      "entropy": 6.333895015716553,
      "epoch": 0.11527377521613832,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999958626333406,
      "loss": 6.1386,
      "mean_token_accuracy": 0.15193476974964143,
      "num_tokens": 2742966.0,
      "step": 1200
    },
    {
      "entropy": 6.265989780426025,
      "epoch": 0.11575408261287223,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999956521142383,
      "loss": 6.0592,
      "mean_token_accuracy": 0.15965109765529634,
      "num_tokens": 2755010.0,
      "step": 1205
    },
    {
      "entropy": 6.39248833656311,
      "epoch": 0.11623439000960614,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999954363714051,
      "loss": 6.0964,
      "mean_token_accuracy": 0.14827171862125396,
      "num_tokens": 2766176.0,
      "step": 1210
    },
    {
      "entropy": 6.379584693908692,
      "epoch": 0.11671469740634005,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999952154048459,
      "loss": 6.2296,
      "mean_token_accuracy": 0.14675042033195496,
      "num_tokens": 2777861.0,
      "step": 1215
    },
    {
      "entropy": 6.266990661621094,
      "epoch": 0.11719500480307397,
      "grad_norm": 1.171875,
      "learning_rate": 0.000499994989214566,
      "loss": 6.0566,
      "mean_token_accuracy": 0.14940764307975768,
      "num_tokens": 2788725.0,
      "step": 1220
    },
    {
      "entropy": 6.2430566310882565,
      "epoch": 0.11767531219980788,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999947578005705,
      "loss": 6.0743,
      "mean_token_accuracy": 0.15087774246931077,
      "num_tokens": 2801613.0,
      "step": 1225
    },
    {
      "entropy": 6.386273336410523,
      "epoch": 0.11815561959654179,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999945211628648,
      "loss": 6.1488,
      "mean_token_accuracy": 0.15173738598823547,
      "num_tokens": 2812474.0,
      "step": 1230
    },
    {
      "entropy": 6.247502326965332,
      "epoch": 0.1186359269932757,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004999942793014544,
      "loss": 6.0647,
      "mean_token_accuracy": 0.15294726118445395,
      "num_tokens": 2823178.0,
      "step": 1235
    },
    {
      "entropy": 6.219358348846436,
      "epoch": 0.11911623439000961,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499994032216345,
      "loss": 6.0531,
      "mean_token_accuracy": 0.1527070477604866,
      "num_tokens": 2836486.0,
      "step": 1240
    },
    {
      "entropy": 6.325845193862915,
      "epoch": 0.11959654178674352,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999937799075422,
      "loss": 6.1189,
      "mean_token_accuracy": 0.15587549209594725,
      "num_tokens": 2847902.0,
      "step": 1245
    },
    {
      "entropy": 6.120101833343506,
      "epoch": 0.12007684918347743,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499993522375052,
      "loss": 5.9993,
      "mean_token_accuracy": 0.15293321758508682,
      "num_tokens": 2859991.0,
      "step": 1250
    },
    {
      "entropy": 6.44623122215271,
      "epoch": 0.12055715658021134,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999932596188802,
      "loss": 6.1907,
      "mean_token_accuracy": 0.1484307639300823,
      "num_tokens": 2870269.0,
      "step": 1255
    },
    {
      "entropy": 6.26564474105835,
      "epoch": 0.12103746397694524,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999929916390331,
      "loss": 6.0708,
      "mean_token_accuracy": 0.1487228438258171,
      "num_tokens": 2882191.0,
      "step": 1260
    },
    {
      "entropy": 6.222062683105468,
      "epoch": 0.12151777137367915,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999927184355169,
      "loss": 6.0777,
      "mean_token_accuracy": 0.14720139130949975,
      "num_tokens": 2892775.0,
      "step": 1265
    },
    {
      "entropy": 6.304709196090698,
      "epoch": 0.12199807877041306,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999924400083377,
      "loss": 6.0619,
      "mean_token_accuracy": 0.15707618668675422,
      "num_tokens": 2904750.0,
      "step": 1270
    },
    {
      "entropy": 6.313355350494385,
      "epoch": 0.12247838616714697,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999921563575022,
      "loss": 6.119,
      "mean_token_accuracy": 0.14971737340092658,
      "num_tokens": 2916150.0,
      "step": 1275
    },
    {
      "entropy": 6.3394777297973635,
      "epoch": 0.12295869356388088,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999918674830169,
      "loss": 6.1047,
      "mean_token_accuracy": 0.14871277660131454,
      "num_tokens": 2928452.0,
      "step": 1280
    },
    {
      "entropy": 6.252659940719605,
      "epoch": 0.12343900096061479,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999915733848886,
      "loss": 6.0908,
      "mean_token_accuracy": 0.14653694331645967,
      "num_tokens": 2940577.0,
      "step": 1285
    },
    {
      "entropy": 6.2738316535949705,
      "epoch": 0.1239193083573487,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499991274063124,
      "loss": 6.0563,
      "mean_token_accuracy": 0.15000832378864287,
      "num_tokens": 2952302.0,
      "step": 1290
    },
    {
      "entropy": 6.277376508712768,
      "epoch": 0.12439961575408261,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004999909695177301,
      "loss": 6.1199,
      "mean_token_accuracy": 0.1537736400961876,
      "num_tokens": 2964611.0,
      "step": 1295
    },
    {
      "entropy": 6.316516351699829,
      "epoch": 0.12487992315081652,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499990659748714,
      "loss": 6.1008,
      "mean_token_accuracy": 0.14772574454545975,
      "num_tokens": 2975668.0,
      "step": 1300
    },
    {
      "entropy": 6.379649114608765,
      "epoch": 0.12536023054755044,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999903447560828,
      "loss": 6.163,
      "mean_token_accuracy": 0.15174921303987504,
      "num_tokens": 2987303.0,
      "step": 1305
    },
    {
      "entropy": 6.3311192989349365,
      "epoch": 0.12584053794428435,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999900245398439,
      "loss": 6.06,
      "mean_token_accuracy": 0.155066779255867,
      "num_tokens": 3000400.0,
      "step": 1310
    },
    {
      "entropy": 6.25769944190979,
      "epoch": 0.12632084534101826,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999896991000047,
      "loss": 5.9963,
      "mean_token_accuracy": 0.15082165524363517,
      "num_tokens": 3012336.0,
      "step": 1315
    },
    {
      "entropy": 6.3025330066680905,
      "epoch": 0.12680115273775217,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004999893684365729,
      "loss": 6.0618,
      "mean_token_accuracy": 0.1516493648290634,
      "num_tokens": 3023004.0,
      "step": 1320
    },
    {
      "entropy": 6.239853048324585,
      "epoch": 0.12728146013448607,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999890325495559,
      "loss": 6.1332,
      "mean_token_accuracy": 0.1481592871248722,
      "num_tokens": 3035147.0,
      "step": 1325
    },
    {
      "entropy": 6.269144773483276,
      "epoch": 0.12776176753121998,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999886914389617,
      "loss": 5.964,
      "mean_token_accuracy": 0.15615130662918092,
      "num_tokens": 3045611.0,
      "step": 1330
    },
    {
      "entropy": 6.211176919937134,
      "epoch": 0.1282420749279539,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999883451047981,
      "loss": 5.9794,
      "mean_token_accuracy": 0.15196620672941208,
      "num_tokens": 3056420.0,
      "step": 1335
    },
    {
      "entropy": 6.195685625076294,
      "epoch": 0.1287223823246878,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999879935470733,
      "loss": 5.963,
      "mean_token_accuracy": 0.15559092015028,
      "num_tokens": 3068770.0,
      "step": 1340
    },
    {
      "entropy": 6.254584407806396,
      "epoch": 0.1292026897214217,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999876367657954,
      "loss": 6.0985,
      "mean_token_accuracy": 0.1450921505689621,
      "num_tokens": 3080806.0,
      "step": 1345
    },
    {
      "entropy": 6.348188734054565,
      "epoch": 0.12968299711815562,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999872747609725,
      "loss": 6.1052,
      "mean_token_accuracy": 0.14815450757741927,
      "num_tokens": 3091769.0,
      "step": 1350
    },
    {
      "entropy": 6.241403007507325,
      "epoch": 0.13016330451488953,
      "grad_norm": 1.25,
      "learning_rate": 0.0004999869075326132,
      "loss": 6.0277,
      "mean_token_accuracy": 0.15045573711395263,
      "num_tokens": 3103121.0,
      "step": 1355
    },
    {
      "entropy": 6.2066021919250485,
      "epoch": 0.13064361191162344,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499986535080726,
      "loss": 6.0157,
      "mean_token_accuracy": 0.15554290562868117,
      "num_tokens": 3115606.0,
      "step": 1360
    },
    {
      "entropy": 6.200456380844116,
      "epoch": 0.13112391930835735,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999861574053196,
      "loss": 5.9027,
      "mean_token_accuracy": 0.16081953197717666,
      "num_tokens": 3127961.0,
      "step": 1365
    },
    {
      "entropy": 6.141843366622925,
      "epoch": 0.13160422670509125,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999857745064027,
      "loss": 5.9219,
      "mean_token_accuracy": 0.16132236272096634,
      "num_tokens": 3138316.0,
      "step": 1370
    },
    {
      "entropy": 6.1357402324676515,
      "epoch": 0.13208453410182516,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499985386383984,
      "loss": 5.9155,
      "mean_token_accuracy": 0.15786689072847365,
      "num_tokens": 3150818.0,
      "step": 1375
    },
    {
      "entropy": 6.233044052124024,
      "epoch": 0.13256484149855907,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999849930380729,
      "loss": 6.0552,
      "mean_token_accuracy": 0.1527769923210144,
      "num_tokens": 3162066.0,
      "step": 1380
    },
    {
      "entropy": 6.123853874206543,
      "epoch": 0.13304514889529298,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999845944686781,
      "loss": 6.0345,
      "mean_token_accuracy": 0.1494033455848694,
      "num_tokens": 3172209.0,
      "step": 1385
    },
    {
      "entropy": 6.176547431945801,
      "epoch": 0.1335254562920269,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999841906758093,
      "loss": 5.8576,
      "mean_token_accuracy": 0.16600174009799956,
      "num_tokens": 3183248.0,
      "step": 1390
    },
    {
      "entropy": 6.142237281799316,
      "epoch": 0.1340057636887608,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999837816594757,
      "loss": 5.9616,
      "mean_token_accuracy": 0.15896387100219728,
      "num_tokens": 3194748.0,
      "step": 1395
    },
    {
      "entropy": 6.1211505889892575,
      "epoch": 0.1344860710854947,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999833674196865,
      "loss": 5.9232,
      "mean_token_accuracy": 0.16622720211744307,
      "num_tokens": 3205669.0,
      "step": 1400
    },
    {
      "entropy": 6.197529077529907,
      "epoch": 0.13496637848222862,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999829479564518,
      "loss": 6.0093,
      "mean_token_accuracy": 0.15054902583360671,
      "num_tokens": 3216035.0,
      "step": 1405
    },
    {
      "entropy": 6.228098821640015,
      "epoch": 0.13544668587896252,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499982523269781,
      "loss": 6.0043,
      "mean_token_accuracy": 0.15448517501354217,
      "num_tokens": 3227192.0,
      "step": 1410
    },
    {
      "entropy": 6.172352313995361,
      "epoch": 0.13592699327569643,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999820933596842,
      "loss": 6.0283,
      "mean_token_accuracy": 0.1561071217060089,
      "num_tokens": 3240237.0,
      "step": 1415
    },
    {
      "entropy": 6.193251514434815,
      "epoch": 0.13640730067243034,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499981658226171,
      "loss": 5.9033,
      "mean_token_accuracy": 0.16419047713279725,
      "num_tokens": 3251963.0,
      "step": 1420
    },
    {
      "entropy": 6.135825729370117,
      "epoch": 0.13688760806916425,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000499981217869252,
      "loss": 6.0291,
      "mean_token_accuracy": 0.16153036952018737,
      "num_tokens": 3263101.0,
      "step": 1425
    },
    {
      "entropy": 6.17513861656189,
      "epoch": 0.1373679154658982,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000499980772288937,
      "loss": 5.9057,
      "mean_token_accuracy": 0.16388555020093917,
      "num_tokens": 3275100.0,
      "step": 1430
    },
    {
      "entropy": 6.167034149169922,
      "epoch": 0.1378482228626321,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999803214852367,
      "loss": 6.0065,
      "mean_token_accuracy": 0.1535780042409897,
      "num_tokens": 3287025.0,
      "step": 1435
    },
    {
      "entropy": 6.230725336074829,
      "epoch": 0.138328530259366,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999798654581613,
      "loss": 6.0077,
      "mean_token_accuracy": 0.15366143360733986,
      "num_tokens": 3299867.0,
      "step": 1440
    },
    {
      "entropy": 6.1359796047210695,
      "epoch": 0.13880883765609991,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999794042077214,
      "loss": 5.9324,
      "mean_token_accuracy": 0.15622105449438095,
      "num_tokens": 3311183.0,
      "step": 1445
    },
    {
      "entropy": 6.170093536376953,
      "epoch": 0.13928914505283382,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999789377339279,
      "loss": 6.0028,
      "mean_token_accuracy": 0.15496657863259317,
      "num_tokens": 3322247.0,
      "step": 1450
    },
    {
      "entropy": 6.195834493637085,
      "epoch": 0.13976945244956773,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999784660367915,
      "loss": 5.9332,
      "mean_token_accuracy": 0.1554928407073021,
      "num_tokens": 3333369.0,
      "step": 1455
    },
    {
      "entropy": 6.079760551452637,
      "epoch": 0.14024975984630164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999779891163231,
      "loss": 5.9575,
      "mean_token_accuracy": 0.15849529206752777,
      "num_tokens": 3345876.0,
      "step": 1460
    },
    {
      "entropy": 6.154807806015015,
      "epoch": 0.14073006724303555,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999775069725339,
      "loss": 5.8498,
      "mean_token_accuracy": 0.1605023294687271,
      "num_tokens": 3357323.0,
      "step": 1465
    },
    {
      "entropy": 6.096273374557495,
      "epoch": 0.14121037463976946,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499977019605435,
      "loss": 5.9311,
      "mean_token_accuracy": 0.15962084680795668,
      "num_tokens": 3367689.0,
      "step": 1470
    },
    {
      "entropy": 6.006128931045533,
      "epoch": 0.14169068203650337,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999765270150378,
      "loss": 5.8952,
      "mean_token_accuracy": 0.16288901418447493,
      "num_tokens": 3379472.0,
      "step": 1475
    },
    {
      "entropy": 6.239159154891968,
      "epoch": 0.14217098943323728,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999760292013536,
      "loss": 5.9533,
      "mean_token_accuracy": 0.15554243996739386,
      "num_tokens": 3390929.0,
      "step": 1480
    },
    {
      "entropy": 6.209478282928467,
      "epoch": 0.14265129682997119,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999755261643941,
      "loss": 5.9401,
      "mean_token_accuracy": 0.16148580312728883,
      "num_tokens": 3401242.0,
      "step": 1485
    },
    {
      "entropy": 6.060628652572632,
      "epoch": 0.1431316042267051,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999750179041709,
      "loss": 5.9322,
      "mean_token_accuracy": 0.1567712590098381,
      "num_tokens": 3411169.0,
      "step": 1490
    },
    {
      "entropy": 6.067643499374389,
      "epoch": 0.143611911623439,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999745044206959,
      "loss": 5.7653,
      "mean_token_accuracy": 0.1653980001807213,
      "num_tokens": 3423265.0,
      "step": 1495
    },
    {
      "entropy": 6.061836051940918,
      "epoch": 0.1440922190201729,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999739857139809,
      "loss": 5.8723,
      "mean_token_accuracy": 0.1658122293651104,
      "num_tokens": 3434793.0,
      "step": 1500
    },
    {
      "entropy": 6.013027095794678,
      "epoch": 0.14457252641690682,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499973461784038,
      "loss": 5.8078,
      "mean_token_accuracy": 0.17317451983690263,
      "num_tokens": 3445732.0,
      "step": 1505
    },
    {
      "entropy": 6.127257823944092,
      "epoch": 0.14505283381364073,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999729326308792,
      "loss": 5.9813,
      "mean_token_accuracy": 0.15522808730602264,
      "num_tokens": 3457090.0,
      "step": 1510
    },
    {
      "entropy": 6.194346618652344,
      "epoch": 0.14553314121037464,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000499972398254517,
      "loss": 5.9726,
      "mean_token_accuracy": 0.15402631014585494,
      "num_tokens": 3468087.0,
      "step": 1515
    },
    {
      "entropy": 6.157139253616333,
      "epoch": 0.14601344860710855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499971858654964,
      "loss": 5.9108,
      "mean_token_accuracy": 0.15916238874197006,
      "num_tokens": 3478820.0,
      "step": 1520
    },
    {
      "entropy": 6.0432126998901365,
      "epoch": 0.14649375600384246,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999713138322321,
      "loss": 5.9387,
      "mean_token_accuracy": 0.15639192685484887,
      "num_tokens": 3489878.0,
      "step": 1525
    },
    {
      "entropy": 6.179038763046265,
      "epoch": 0.14697406340057637,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999707637863346,
      "loss": 5.9227,
      "mean_token_accuracy": 0.15898001939058304,
      "num_tokens": 3500944.0,
      "step": 1530
    },
    {
      "entropy": 6.006661891937256,
      "epoch": 0.14745437079731027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999702085172838,
      "loss": 5.8949,
      "mean_token_accuracy": 0.166755573451519,
      "num_tokens": 3511383.0,
      "step": 1535
    },
    {
      "entropy": 6.193773031234741,
      "epoch": 0.14793467819404418,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999696480250929,
      "loss": 6.0009,
      "mean_token_accuracy": 0.1563890740275383,
      "num_tokens": 3523300.0,
      "step": 1540
    },
    {
      "entropy": 6.137639045715332,
      "epoch": 0.1484149855907781,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004999690823097747,
      "loss": 5.904,
      "mean_token_accuracy": 0.15553956106305122,
      "num_tokens": 3534371.0,
      "step": 1545
    },
    {
      "entropy": 6.11188440322876,
      "epoch": 0.148895292987512,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999685113713426,
      "loss": 5.8848,
      "mean_token_accuracy": 0.1581243097782135,
      "num_tokens": 3544847.0,
      "step": 1550
    },
    {
      "entropy": 6.084521484375,
      "epoch": 0.1493756003842459,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999679352098096,
      "loss": 5.8683,
      "mean_token_accuracy": 0.1633513331413269,
      "num_tokens": 3555859.0,
      "step": 1555
    },
    {
      "entropy": 6.081562280654907,
      "epoch": 0.14985590778097982,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999673538251891,
      "loss": 5.8724,
      "mean_token_accuracy": 0.16277971267700195,
      "num_tokens": 3568283.0,
      "step": 1560
    },
    {
      "entropy": 6.040332508087158,
      "epoch": 0.15033621517771373,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999667672174947,
      "loss": 5.9535,
      "mean_token_accuracy": 0.15841858685016633,
      "num_tokens": 3581442.0,
      "step": 1565
    },
    {
      "entropy": 6.214689922332764,
      "epoch": 0.15081652257444764,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00049996617538674,
      "loss": 5.999,
      "mean_token_accuracy": 0.15400405079126359,
      "num_tokens": 3594055.0,
      "step": 1570
    },
    {
      "entropy": 6.139510774612427,
      "epoch": 0.15129682997118155,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999655783329386,
      "loss": 5.9564,
      "mean_token_accuracy": 0.1526302471756935,
      "num_tokens": 3605952.0,
      "step": 1575
    },
    {
      "entropy": 6.104340982437134,
      "epoch": 0.15177713736791545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999649760561046,
      "loss": 5.9762,
      "mean_token_accuracy": 0.15697605460882186,
      "num_tokens": 3618544.0,
      "step": 1580
    },
    {
      "entropy": 6.139195299148559,
      "epoch": 0.15225744476464936,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999643685562519,
      "loss": 5.928,
      "mean_token_accuracy": 0.16384828090667725,
      "num_tokens": 3630445.0,
      "step": 1585
    },
    {
      "entropy": 6.140196180343628,
      "epoch": 0.15273775216138327,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999637558333945,
      "loss": 5.9277,
      "mean_token_accuracy": 0.162806536257267,
      "num_tokens": 3642516.0,
      "step": 1590
    },
    {
      "entropy": 6.032298421859741,
      "epoch": 0.15321805955811718,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999631378875467,
      "loss": 5.8473,
      "mean_token_accuracy": 0.1700352057814598,
      "num_tokens": 3654425.0,
      "step": 1595
    },
    {
      "entropy": 6.052908039093017,
      "epoch": 0.15369836695485112,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999625147187228,
      "loss": 5.8659,
      "mean_token_accuracy": 0.1607646018266678,
      "num_tokens": 3666521.0,
      "step": 1600
    },
    {
      "entropy": 6.144074392318726,
      "epoch": 0.15417867435158503,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999618863269373,
      "loss": 5.9113,
      "mean_token_accuracy": 0.15713147521018983,
      "num_tokens": 3679121.0,
      "step": 1605
    },
    {
      "entropy": 6.08037257194519,
      "epoch": 0.15465898174831894,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999612527122049,
      "loss": 5.9215,
      "mean_token_accuracy": 0.15551023483276366,
      "num_tokens": 3691095.0,
      "step": 1610
    },
    {
      "entropy": 6.048700380325317,
      "epoch": 0.15513928914505284,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999606138745402,
      "loss": 5.8863,
      "mean_token_accuracy": 0.16148294806480407,
      "num_tokens": 3703426.0,
      "step": 1615
    },
    {
      "entropy": 6.160890388488769,
      "epoch": 0.15561959654178675,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999599698139581,
      "loss": 5.9506,
      "mean_token_accuracy": 0.16333511769771575,
      "num_tokens": 3715429.0,
      "step": 1620
    },
    {
      "entropy": 6.134197616577149,
      "epoch": 0.15609990393852066,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999593205304734,
      "loss": 5.9302,
      "mean_token_accuracy": 0.1580351047217846,
      "num_tokens": 3726327.0,
      "step": 1625
    },
    {
      "entropy": 6.134684467315674,
      "epoch": 0.15658021133525457,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999586660241012,
      "loss": 5.9004,
      "mean_token_accuracy": 0.15458485931158067,
      "num_tokens": 3736818.0,
      "step": 1630
    },
    {
      "entropy": 6.102252435684204,
      "epoch": 0.15706051873198848,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999580062948569,
      "loss": 5.9023,
      "mean_token_accuracy": 0.16163242906332015,
      "num_tokens": 3747776.0,
      "step": 1635
    },
    {
      "entropy": 5.938621568679809,
      "epoch": 0.1575408261287224,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999573413427556,
      "loss": 5.7522,
      "mean_token_accuracy": 0.16386649161577224,
      "num_tokens": 3758990.0,
      "step": 1640
    },
    {
      "entropy": 6.073081064224243,
      "epoch": 0.1580211335254563,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999566711678128,
      "loss": 5.8335,
      "mean_token_accuracy": 0.1618332728743553,
      "num_tokens": 3769686.0,
      "step": 1645
    },
    {
      "entropy": 6.034470748901367,
      "epoch": 0.1585014409221902,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999559957700442,
      "loss": 5.8842,
      "mean_token_accuracy": 0.16474503725767137,
      "num_tokens": 3781815.0,
      "step": 1650
    },
    {
      "entropy": 6.107155132293701,
      "epoch": 0.15898174831892412,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999553151494653,
      "loss": 5.9224,
      "mean_token_accuracy": 0.16111959964036943,
      "num_tokens": 3793392.0,
      "step": 1655
    },
    {
      "entropy": 6.020937395095825,
      "epoch": 0.15946205571565802,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999546293060919,
      "loss": 5.8622,
      "mean_token_accuracy": 0.15857846289873123,
      "num_tokens": 3804974.0,
      "step": 1660
    },
    {
      "entropy": 5.980496025085449,
      "epoch": 0.15994236311239193,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049995393823994,
      "loss": 5.7399,
      "mean_token_accuracy": 0.1712213933467865,
      "num_tokens": 3817166.0,
      "step": 1665
    },
    {
      "entropy": 5.985702800750732,
      "epoch": 0.16042267050912584,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999532419510255,
      "loss": 5.8584,
      "mean_token_accuracy": 0.16114549189805985,
      "num_tokens": 3828151.0,
      "step": 1670
    },
    {
      "entropy": 6.079834032058716,
      "epoch": 0.16090297790585975,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499952540439365,
      "loss": 5.8566,
      "mean_token_accuracy": 0.16157617270946503,
      "num_tokens": 3839439.0,
      "step": 1675
    },
    {
      "entropy": 6.109479236602783,
      "epoch": 0.16138328530259366,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999518337049743,
      "loss": 5.9183,
      "mean_token_accuracy": 0.15559976994991304,
      "num_tokens": 3851694.0,
      "step": 1680
    },
    {
      "entropy": 5.999596118927002,
      "epoch": 0.16186359269932757,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049995112174787,
      "loss": 5.8943,
      "mean_token_accuracy": 0.15952107906341553,
      "num_tokens": 3863593.0,
      "step": 1685
    },
    {
      "entropy": 5.972264003753662,
      "epoch": 0.16234390009606148,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999504045680687,
      "loss": 5.8039,
      "mean_token_accuracy": 0.16952102333307267,
      "num_tokens": 3874588.0,
      "step": 1690
    },
    {
      "entropy": 6.094362878799439,
      "epoch": 0.1628242074927954,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999496821655869,
      "loss": 5.8981,
      "mean_token_accuracy": 0.15856973826885223,
      "num_tokens": 3884662.0,
      "step": 1695
    },
    {
      "entropy": 6.140773725509644,
      "epoch": 0.1633045148895293,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999489545404414,
      "loss": 5.9891,
      "mean_token_accuracy": 0.15623455494642258,
      "num_tokens": 3896569.0,
      "step": 1700
    },
    {
      "entropy": 6.140805101394653,
      "epoch": 0.1637848222862632,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999482216926493,
      "loss": 5.8462,
      "mean_token_accuracy": 0.16176871806383133,
      "num_tokens": 3907691.0,
      "step": 1705
    },
    {
      "entropy": 6.034537076950073,
      "epoch": 0.1642651296829971,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999474836222273,
      "loss": 5.8498,
      "mean_token_accuracy": 0.16770250350236893,
      "num_tokens": 3918794.0,
      "step": 1710
    },
    {
      "entropy": 6.016002893447876,
      "epoch": 0.16474543707973102,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999467403291928,
      "loss": 5.8424,
      "mean_token_accuracy": 0.17401330322027206,
      "num_tokens": 3929773.0,
      "step": 1715
    },
    {
      "entropy": 6.039723968505859,
      "epoch": 0.16522574447646493,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999459918135628,
      "loss": 5.8626,
      "mean_token_accuracy": 0.16028911992907524,
      "num_tokens": 3940264.0,
      "step": 1720
    },
    {
      "entropy": 5.988501930236817,
      "epoch": 0.16570605187319884,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499945238075355,
      "loss": 5.7585,
      "mean_token_accuracy": 0.16902887523174287,
      "num_tokens": 3951500.0,
      "step": 1725
    },
    {
      "entropy": 6.034324693679809,
      "epoch": 0.16618635926993275,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999444791145865,
      "loss": 5.8427,
      "mean_token_accuracy": 0.16535028964281082,
      "num_tokens": 3963580.0,
      "step": 1730
    },
    {
      "entropy": 5.986796712875366,
      "epoch": 0.16666666666666666,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999437149312754,
      "loss": 5.7757,
      "mean_token_accuracy": 0.17422386556863784,
      "num_tokens": 3975994.0,
      "step": 1735
    },
    {
      "entropy": 5.9908356189727785,
      "epoch": 0.16714697406340057,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499942945525439,
      "loss": 5.7922,
      "mean_token_accuracy": 0.1610643297433853,
      "num_tokens": 3987897.0,
      "step": 1740
    },
    {
      "entropy": 6.048891305923462,
      "epoch": 0.16762728146013448,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999421708970954,
      "loss": 5.9403,
      "mean_token_accuracy": 0.1578907661139965,
      "num_tokens": 3999829.0,
      "step": 1745
    },
    {
      "entropy": 6.001609849929809,
      "epoch": 0.16810758885686838,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999413910462625,
      "loss": 5.7895,
      "mean_token_accuracy": 0.16723742038011552,
      "num_tokens": 4010882.0,
      "step": 1750
    },
    {
      "entropy": 5.978082466125488,
      "epoch": 0.1685878962536023,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999406059729586,
      "loss": 5.7717,
      "mean_token_accuracy": 0.16948328018188477,
      "num_tokens": 4021423.0,
      "step": 1755
    },
    {
      "entropy": 6.093415307998657,
      "epoch": 0.1690682036503362,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999398156772016,
      "loss": 5.9183,
      "mean_token_accuracy": 0.15228435546159744,
      "num_tokens": 4033590.0,
      "step": 1760
    },
    {
      "entropy": 5.927422332763672,
      "epoch": 0.16954851104707014,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00049993902015901,
      "loss": 5.7775,
      "mean_token_accuracy": 0.16497890055179595,
      "num_tokens": 4043978.0,
      "step": 1765
    },
    {
      "entropy": 6.080128145217896,
      "epoch": 0.17002881844380405,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999382194184023,
      "loss": 5.9011,
      "mean_token_accuracy": 0.16162478625774385,
      "num_tokens": 4054513.0,
      "step": 1770
    },
    {
      "entropy": 6.0692665576934814,
      "epoch": 0.17050912584053796,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999374134553972,
      "loss": 5.8467,
      "mean_token_accuracy": 0.16818321645259857,
      "num_tokens": 4066019.0,
      "step": 1775
    },
    {
      "entropy": 5.963143253326416,
      "epoch": 0.17098943323727187,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999366022700131,
      "loss": 5.8214,
      "mean_token_accuracy": 0.16922373026609422,
      "num_tokens": 4077688.0,
      "step": 1780
    },
    {
      "entropy": 6.082214260101319,
      "epoch": 0.17146974063400577,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999357858622691,
      "loss": 5.8939,
      "mean_token_accuracy": 0.1704793967306614,
      "num_tokens": 4089803.0,
      "step": 1785
    },
    {
      "entropy": 6.019527435302734,
      "epoch": 0.17195004803073968,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999349642321842,
      "loss": 5.8323,
      "mean_token_accuracy": 0.16811123341321946,
      "num_tokens": 4101969.0,
      "step": 1790
    },
    {
      "entropy": 5.995560216903686,
      "epoch": 0.1724303554274736,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999341373797772,
      "loss": 5.8181,
      "mean_token_accuracy": 0.16215983778238297,
      "num_tokens": 4113567.0,
      "step": 1795
    },
    {
      "entropy": 5.987828254699707,
      "epoch": 0.1729106628242075,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999333053050675,
      "loss": 5.7861,
      "mean_token_accuracy": 0.1671761840581894,
      "num_tokens": 4125191.0,
      "step": 1800
    },
    {
      "entropy": 6.054584407806397,
      "epoch": 0.1733909702209414,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004999324680080744,
      "loss": 5.8297,
      "mean_token_accuracy": 0.167048180103302,
      "num_tokens": 4135050.0,
      "step": 1805
    },
    {
      "entropy": 5.962006425857544,
      "epoch": 0.17387127761767532,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999316254888172,
      "loss": 5.8864,
      "mean_token_accuracy": 0.16525756269693376,
      "num_tokens": 4146874.0,
      "step": 1810
    },
    {
      "entropy": 6.027948713302612,
      "epoch": 0.17435158501440923,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999307777473157,
      "loss": 5.8119,
      "mean_token_accuracy": 0.16480804234743118,
      "num_tokens": 4158118.0,
      "step": 1815
    },
    {
      "entropy": 5.980266284942627,
      "epoch": 0.17483189241114314,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999299247835893,
      "loss": 5.7898,
      "mean_token_accuracy": 0.17785038501024247,
      "num_tokens": 4169035.0,
      "step": 1820
    },
    {
      "entropy": 5.911714839935303,
      "epoch": 0.17531219980787704,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499929066597658,
      "loss": 5.767,
      "mean_token_accuracy": 0.17150768488645554,
      "num_tokens": 4180314.0,
      "step": 1825
    },
    {
      "entropy": 6.06360354423523,
      "epoch": 0.17579250720461095,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999282031895418,
      "loss": 5.8574,
      "mean_token_accuracy": 0.1665028989315033,
      "num_tokens": 4192238.0,
      "step": 1830
    },
    {
      "entropy": 5.918634033203125,
      "epoch": 0.17627281460134486,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999273345592604,
      "loss": 5.7864,
      "mean_token_accuracy": 0.16353099048137665,
      "num_tokens": 4203346.0,
      "step": 1835
    },
    {
      "entropy": 5.99358344078064,
      "epoch": 0.17675312199807877,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999264607068343,
      "loss": 5.8248,
      "mean_token_accuracy": 0.16922380030155182,
      "num_tokens": 4213763.0,
      "step": 1840
    },
    {
      "entropy": 5.949887943267822,
      "epoch": 0.17723342939481268,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999255816322837,
      "loss": 5.7987,
      "mean_token_accuracy": 0.1714692607522011,
      "num_tokens": 4225553.0,
      "step": 1845
    },
    {
      "entropy": 6.06374306678772,
      "epoch": 0.1777137367915466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499924697335629,
      "loss": 5.7171,
      "mean_token_accuracy": 0.17483123242855073,
      "num_tokens": 4236058.0,
      "step": 1850
    },
    {
      "entropy": 5.883856916427613,
      "epoch": 0.1781940441882805,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999238078168906,
      "loss": 5.7983,
      "mean_token_accuracy": 0.17132625132799148,
      "num_tokens": 4248299.0,
      "step": 1855
    },
    {
      "entropy": 5.987727499008178,
      "epoch": 0.1786743515850144,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999229130760894,
      "loss": 5.7295,
      "mean_token_accuracy": 0.16837194561958313,
      "num_tokens": 4259704.0,
      "step": 1860
    },
    {
      "entropy": 5.883204078674316,
      "epoch": 0.17915465898174832,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499922013113246,
      "loss": 5.6159,
      "mean_token_accuracy": 0.183360692858696,
      "num_tokens": 4270480.0,
      "step": 1865
    },
    {
      "entropy": 5.932878494262695,
      "epoch": 0.17963496637848222,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999211079283814,
      "loss": 5.8661,
      "mean_token_accuracy": 0.16950789242982864,
      "num_tokens": 4282104.0,
      "step": 1870
    },
    {
      "entropy": 6.039040184020996,
      "epoch": 0.18011527377521613,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999201975215164,
      "loss": 5.8454,
      "mean_token_accuracy": 0.16162874400615693,
      "num_tokens": 4294251.0,
      "step": 1875
    },
    {
      "entropy": 5.932565784454345,
      "epoch": 0.18059558117195004,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999192818926725,
      "loss": 5.7256,
      "mean_token_accuracy": 0.16642534136772155,
      "num_tokens": 4305569.0,
      "step": 1880
    },
    {
      "entropy": 6.010633945465088,
      "epoch": 0.18107588856868395,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999183610418706,
      "loss": 5.856,
      "mean_token_accuracy": 0.16204970329999924,
      "num_tokens": 4317845.0,
      "step": 1885
    },
    {
      "entropy": 5.937364482879639,
      "epoch": 0.18155619596541786,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999174349691322,
      "loss": 5.7241,
      "mean_token_accuracy": 0.16895480155944825,
      "num_tokens": 4329987.0,
      "step": 1890
    },
    {
      "entropy": 5.916909265518188,
      "epoch": 0.18203650336215177,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999165036744788,
      "loss": 5.7336,
      "mean_token_accuracy": 0.16701279431581498,
      "num_tokens": 4341628.0,
      "step": 1895
    },
    {
      "entropy": 5.97577314376831,
      "epoch": 0.18251681075888568,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999155671579322,
      "loss": 5.8141,
      "mean_token_accuracy": 0.16247764229774475,
      "num_tokens": 4352379.0,
      "step": 1900
    },
    {
      "entropy": 5.888210439682007,
      "epoch": 0.1829971181556196,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499914625419514,
      "loss": 5.7384,
      "mean_token_accuracy": 0.17271625846624375,
      "num_tokens": 4364800.0,
      "step": 1905
    },
    {
      "entropy": 5.934367513656616,
      "epoch": 0.1834774255523535,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999136784592459,
      "loss": 5.7544,
      "mean_token_accuracy": 0.16907702684402465,
      "num_tokens": 4376048.0,
      "step": 1910
    },
    {
      "entropy": 5.916653251647949,
      "epoch": 0.1839577329490874,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999127262771502,
      "loss": 5.7474,
      "mean_token_accuracy": 0.16713621020317077,
      "num_tokens": 4388072.0,
      "step": 1915
    },
    {
      "entropy": 6.049605655670166,
      "epoch": 0.1844380403458213,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999117688732487,
      "loss": 5.8408,
      "mean_token_accuracy": 0.16282255798578263,
      "num_tokens": 4399843.0,
      "step": 1920
    },
    {
      "entropy": 5.854936361312866,
      "epoch": 0.18491834774255522,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999108062475638,
      "loss": 5.692,
      "mean_token_accuracy": 0.17128440588712693,
      "num_tokens": 4411373.0,
      "step": 1925
    },
    {
      "entropy": 5.888366031646728,
      "epoch": 0.18539865513928913,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499909838400118,
      "loss": 5.6829,
      "mean_token_accuracy": 0.1735598623752594,
      "num_tokens": 4421857.0,
      "step": 1930
    },
    {
      "entropy": 5.947378730773925,
      "epoch": 0.18587896253602307,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999088653309334,
      "loss": 5.7819,
      "mean_token_accuracy": 0.1680931806564331,
      "num_tokens": 4432728.0,
      "step": 1935
    },
    {
      "entropy": 5.886118745803833,
      "epoch": 0.18635926993275698,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999078870400329,
      "loss": 5.714,
      "mean_token_accuracy": 0.1756661355495453,
      "num_tokens": 4444683.0,
      "step": 1940
    },
    {
      "entropy": 5.994152688980103,
      "epoch": 0.18683957732949089,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999069035274391,
      "loss": 5.8381,
      "mean_token_accuracy": 0.16196750849485397,
      "num_tokens": 4456961.0,
      "step": 1945
    },
    {
      "entropy": 5.912719821929931,
      "epoch": 0.1873198847262248,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999059147931747,
      "loss": 5.6738,
      "mean_token_accuracy": 0.17798083573579787,
      "num_tokens": 4468424.0,
      "step": 1950
    },
    {
      "entropy": 5.929690361022949,
      "epoch": 0.1878001921229587,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999049208372629,
      "loss": 5.8701,
      "mean_token_accuracy": 0.15679999366402625,
      "num_tokens": 4479813.0,
      "step": 1955
    },
    {
      "entropy": 6.098432826995849,
      "epoch": 0.1882804995196926,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999039216597267,
      "loss": 5.8826,
      "mean_token_accuracy": 0.16513065099716187,
      "num_tokens": 4491172.0,
      "step": 1960
    },
    {
      "entropy": 5.891456174850464,
      "epoch": 0.18876080691642652,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999029172605892,
      "loss": 5.7655,
      "mean_token_accuracy": 0.17174434214830397,
      "num_tokens": 4503063.0,
      "step": 1965
    },
    {
      "entropy": 6.019789409637451,
      "epoch": 0.18924111431316043,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999019076398738,
      "loss": 5.8443,
      "mean_token_accuracy": 0.15457844734191895,
      "num_tokens": 4514188.0,
      "step": 1970
    },
    {
      "entropy": 5.98090615272522,
      "epoch": 0.18972142170989434,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499900892797604,
      "loss": 5.7495,
      "mean_token_accuracy": 0.17291804850101472,
      "num_tokens": 4525293.0,
      "step": 1975
    },
    {
      "entropy": 5.9353631973266605,
      "epoch": 0.19020172910662825,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998998727338031,
      "loss": 5.8315,
      "mean_token_accuracy": 0.16792338639497756,
      "num_tokens": 4536589.0,
      "step": 1980
    },
    {
      "entropy": 5.87034125328064,
      "epoch": 0.19068203650336216,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998988474484952,
      "loss": 5.5689,
      "mean_token_accuracy": 0.19097184985876084,
      "num_tokens": 4547594.0,
      "step": 1985
    },
    {
      "entropy": 5.888665723800659,
      "epoch": 0.19116234390009607,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998978169417038,
      "loss": 5.7814,
      "mean_token_accuracy": 0.173173625767231,
      "num_tokens": 4559850.0,
      "step": 1990
    },
    {
      "entropy": 5.935269927978515,
      "epoch": 0.19164265129682997,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998967812134529,
      "loss": 5.7343,
      "mean_token_accuracy": 0.16940659731626512,
      "num_tokens": 4570727.0,
      "step": 1995
    },
    {
      "entropy": 5.766289901733399,
      "epoch": 0.19212295869356388,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998957402637664,
      "loss": 5.6639,
      "mean_token_accuracy": 0.17158041000366211,
      "num_tokens": 4582248.0,
      "step": 2000
    },
    {
      "entropy": 5.963186931610108,
      "epoch": 0.1926032660902978,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998946940926687,
      "loss": 5.7005,
      "mean_token_accuracy": 0.16607439517974854,
      "num_tokens": 4592604.0,
      "step": 2005
    },
    {
      "entropy": 5.888526773452758,
      "epoch": 0.1930835734870317,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499893642700184,
      "loss": 5.7435,
      "mean_token_accuracy": 0.1702524498105049,
      "num_tokens": 4604398.0,
      "step": 2010
    },
    {
      "entropy": 5.91964282989502,
      "epoch": 0.1935638808837656,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998925860863368,
      "loss": 5.8039,
      "mean_token_accuracy": 0.16930960565805436,
      "num_tokens": 4616434.0,
      "step": 2015
    },
    {
      "entropy": 5.981127023696899,
      "epoch": 0.19404418828049952,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998915242511516,
      "loss": 5.7597,
      "mean_token_accuracy": 0.17877342998981477,
      "num_tokens": 4627577.0,
      "step": 2020
    },
    {
      "entropy": 5.960260248184204,
      "epoch": 0.19452449567723343,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998904571946528,
      "loss": 5.8227,
      "mean_token_accuracy": 0.16877581179141998,
      "num_tokens": 4639698.0,
      "step": 2025
    },
    {
      "entropy": 6.018915462493896,
      "epoch": 0.19500480307396734,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998893849168655,
      "loss": 5.8322,
      "mean_token_accuracy": 0.16535419821739197,
      "num_tokens": 4650643.0,
      "step": 2030
    },
    {
      "entropy": 5.94498839378357,
      "epoch": 0.19548511047070125,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998883074178144,
      "loss": 5.7511,
      "mean_token_accuracy": 0.17331523150205613,
      "num_tokens": 4662897.0,
      "step": 2035
    },
    {
      "entropy": 5.939677953720093,
      "epoch": 0.19596541786743515,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998872246975247,
      "loss": 5.8446,
      "mean_token_accuracy": 0.1710888773202896,
      "num_tokens": 4673701.0,
      "step": 2040
    },
    {
      "entropy": 6.051514387130737,
      "epoch": 0.19644572526416906,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998861367560213,
      "loss": 5.7976,
      "mean_token_accuracy": 0.17028483599424363,
      "num_tokens": 4685873.0,
      "step": 2045
    },
    {
      "entropy": 5.855360126495361,
      "epoch": 0.19692603266090297,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998850435933296,
      "loss": 5.6799,
      "mean_token_accuracy": 0.171894171833992,
      "num_tokens": 4697179.0,
      "step": 2050
    },
    {
      "entropy": 5.93500165939331,
      "epoch": 0.19740634005763688,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998839452094749,
      "loss": 5.7239,
      "mean_token_accuracy": 0.17378782033920287,
      "num_tokens": 4707752.0,
      "step": 2055
    },
    {
      "entropy": 5.785550880432129,
      "epoch": 0.1978866474543708,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998828416044829,
      "loss": 5.5828,
      "mean_token_accuracy": 0.17932410091161727,
      "num_tokens": 4718413.0,
      "step": 2060
    },
    {
      "entropy": 5.901092576980591,
      "epoch": 0.1983669548511047,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499881732778379,
      "loss": 5.7755,
      "mean_token_accuracy": 0.167689011991024,
      "num_tokens": 4730033.0,
      "step": 2065
    },
    {
      "entropy": 5.849337959289551,
      "epoch": 0.1988472622478386,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499880618731189,
      "loss": 5.632,
      "mean_token_accuracy": 0.17377811521291733,
      "num_tokens": 4742084.0,
      "step": 2070
    },
    {
      "entropy": 5.942147588729858,
      "epoch": 0.19932756964457252,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998794994629388,
      "loss": 5.8568,
      "mean_token_accuracy": 0.16341482400894164,
      "num_tokens": 4753885.0,
      "step": 2075
    },
    {
      "entropy": 5.894834518432617,
      "epoch": 0.19980787704130643,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998783749736545,
      "loss": 5.7098,
      "mean_token_accuracy": 0.1734362706542015,
      "num_tokens": 4765686.0,
      "step": 2080
    },
    {
      "entropy": 5.933993625640869,
      "epoch": 0.20028818443804033,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998772452633619,
      "loss": 5.7384,
      "mean_token_accuracy": 0.16806679219007492,
      "num_tokens": 4777157.0,
      "step": 2085
    },
    {
      "entropy": 5.881355714797974,
      "epoch": 0.20076849183477424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998761103320876,
      "loss": 5.6823,
      "mean_token_accuracy": 0.1787089616060257,
      "num_tokens": 4788583.0,
      "step": 2090
    },
    {
      "entropy": 5.965854835510254,
      "epoch": 0.20124879923150815,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998749701798577,
      "loss": 5.7908,
      "mean_token_accuracy": 0.1650993473827839,
      "num_tokens": 4800749.0,
      "step": 2095
    },
    {
      "entropy": 5.834138011932373,
      "epoch": 0.2017291066282421,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998738248066986,
      "loss": 5.7169,
      "mean_token_accuracy": 0.17363064736127853,
      "num_tokens": 4812488.0,
      "step": 2100
    },
    {
      "entropy": 5.952000665664673,
      "epoch": 0.202209414024976,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998726742126372,
      "loss": 5.6942,
      "mean_token_accuracy": 0.16981008946895598,
      "num_tokens": 4823495.0,
      "step": 2105
    },
    {
      "entropy": 5.821967363357544,
      "epoch": 0.2026897214217099,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998715183976999,
      "loss": 5.7474,
      "mean_token_accuracy": 0.17119863927364348,
      "num_tokens": 4834450.0,
      "step": 2110
    },
    {
      "entropy": 5.896064186096192,
      "epoch": 0.20317002881844382,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998703573619137,
      "loss": 5.653,
      "mean_token_accuracy": 0.18205936402082443,
      "num_tokens": 4846826.0,
      "step": 2115
    },
    {
      "entropy": 5.961601400375367,
      "epoch": 0.20365033621517772,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998691911053056,
      "loss": 5.8431,
      "mean_token_accuracy": 0.16039666682481765,
      "num_tokens": 4859668.0,
      "step": 2120
    },
    {
      "entropy": 5.8840587615966795,
      "epoch": 0.20413064361191163,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998680196279026,
      "loss": 5.7134,
      "mean_token_accuracy": 0.17416077703237534,
      "num_tokens": 4871727.0,
      "step": 2125
    },
    {
      "entropy": 5.968330240249633,
      "epoch": 0.20461095100864554,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998668429297319,
      "loss": 5.78,
      "mean_token_accuracy": 0.1742082878947258,
      "num_tokens": 4882191.0,
      "step": 2130
    },
    {
      "entropy": 5.897537279129028,
      "epoch": 0.20509125840537945,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998656610108208,
      "loss": 5.7018,
      "mean_token_accuracy": 0.1743673115968704,
      "num_tokens": 4892416.0,
      "step": 2135
    },
    {
      "entropy": 5.846418809890747,
      "epoch": 0.20557156580211336,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998644738711969,
      "loss": 5.6797,
      "mean_token_accuracy": 0.1706405311822891,
      "num_tokens": 4903572.0,
      "step": 2140
    },
    {
      "entropy": 5.968627214431763,
      "epoch": 0.20605187319884727,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998632815108874,
      "loss": 5.7787,
      "mean_token_accuracy": 0.16432214826345443,
      "num_tokens": 4915417.0,
      "step": 2145
    },
    {
      "entropy": 5.875264406204224,
      "epoch": 0.20653218059558118,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998620839299203,
      "loss": 5.6596,
      "mean_token_accuracy": 0.17557333707809447,
      "num_tokens": 4926943.0,
      "step": 2150
    },
    {
      "entropy": 5.82595853805542,
      "epoch": 0.2070124879923151,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998608811283233,
      "loss": 5.6216,
      "mean_token_accuracy": 0.1786582961678505,
      "num_tokens": 4937724.0,
      "step": 2155
    },
    {
      "entropy": 5.967111301422119,
      "epoch": 0.207492795389049,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998596731061244,
      "loss": 5.7773,
      "mean_token_accuracy": 0.16382010281085968,
      "num_tokens": 4949970.0,
      "step": 2160
    },
    {
      "entropy": 5.899398708343506,
      "epoch": 0.2079731027857829,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998584598633516,
      "loss": 5.7756,
      "mean_token_accuracy": 0.1687404468655586,
      "num_tokens": 4961389.0,
      "step": 2165
    },
    {
      "entropy": 5.956924343109131,
      "epoch": 0.2084534101825168,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998572414000329,
      "loss": 5.8212,
      "mean_token_accuracy": 0.16519832164049147,
      "num_tokens": 4973888.0,
      "step": 2170
    },
    {
      "entropy": 5.873337078094482,
      "epoch": 0.20893371757925072,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998560177161969,
      "loss": 5.7713,
      "mean_token_accuracy": 0.1622347429394722,
      "num_tokens": 4985423.0,
      "step": 2175
    },
    {
      "entropy": 5.875820159912109,
      "epoch": 0.20941402497598463,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998547888118718,
      "loss": 5.7277,
      "mean_token_accuracy": 0.16388818621635437,
      "num_tokens": 4997711.0,
      "step": 2180
    },
    {
      "entropy": 5.944964838027954,
      "epoch": 0.20989433237271854,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998535546870862,
      "loss": 5.7587,
      "mean_token_accuracy": 0.17094310671091079,
      "num_tokens": 5009633.0,
      "step": 2185
    },
    {
      "entropy": 5.89435224533081,
      "epoch": 0.21037463976945245,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998523153418687,
      "loss": 5.6898,
      "mean_token_accuracy": 0.17721298336982727,
      "num_tokens": 5021523.0,
      "step": 2190
    },
    {
      "entropy": 5.954970121383667,
      "epoch": 0.21085494716618636,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998510707762481,
      "loss": 5.7963,
      "mean_token_accuracy": 0.1734875962138176,
      "num_tokens": 5033513.0,
      "step": 2195
    },
    {
      "entropy": 6.01756477355957,
      "epoch": 0.21133525456292027,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998498209902533,
      "loss": 5.7897,
      "mean_token_accuracy": 0.16785951554775239,
      "num_tokens": 5047055.0,
      "step": 2200
    },
    {
      "entropy": 5.842679691314697,
      "epoch": 0.21181556195965417,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998485659839134,
      "loss": 5.6612,
      "mean_token_accuracy": 0.17763629406690598,
      "num_tokens": 5057613.0,
      "step": 2205
    },
    {
      "entropy": 5.941239261627198,
      "epoch": 0.21229586935638808,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998473057572575,
      "loss": 5.7791,
      "mean_token_accuracy": 0.170269151031971,
      "num_tokens": 5068886.0,
      "step": 2210
    },
    {
      "entropy": 5.8616218090057375,
      "epoch": 0.212776176753122,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998460403103146,
      "loss": 5.7561,
      "mean_token_accuracy": 0.16782673597335815,
      "num_tokens": 5079978.0,
      "step": 2215
    },
    {
      "entropy": 5.881478404998779,
      "epoch": 0.2132564841498559,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998447696431146,
      "loss": 5.7203,
      "mean_token_accuracy": 0.1720557138323784,
      "num_tokens": 5091021.0,
      "step": 2220
    },
    {
      "entropy": 5.84704737663269,
      "epoch": 0.2137367915465898,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998434937556865,
      "loss": 5.6118,
      "mean_token_accuracy": 0.18117671012878417,
      "num_tokens": 5101483.0,
      "step": 2225
    },
    {
      "entropy": 5.8953385829925535,
      "epoch": 0.21421709894332372,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998422126480602,
      "loss": 5.7518,
      "mean_token_accuracy": 0.16402507573366165,
      "num_tokens": 5113116.0,
      "step": 2230
    },
    {
      "entropy": 5.954845809936524,
      "epoch": 0.21469740634005763,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998409263202653,
      "loss": 5.6943,
      "mean_token_accuracy": 0.17064779102802277,
      "num_tokens": 5124824.0,
      "step": 2235
    },
    {
      "entropy": 5.77202959060669,
      "epoch": 0.21517771373679154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998396347723318,
      "loss": 5.6247,
      "mean_token_accuracy": 0.1712438568472862,
      "num_tokens": 5137567.0,
      "step": 2240
    },
    {
      "entropy": 5.833844184875488,
      "epoch": 0.21565802113352545,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998383380042895,
      "loss": 5.6923,
      "mean_token_accuracy": 0.17277579456567765,
      "num_tokens": 5149016.0,
      "step": 2245
    },
    {
      "entropy": 5.808039903640747,
      "epoch": 0.21613832853025935,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998370360161688,
      "loss": 5.5749,
      "mean_token_accuracy": 0.1784022033214569,
      "num_tokens": 5160356.0,
      "step": 2250
    },
    {
      "entropy": 5.881575202941894,
      "epoch": 0.21661863592699326,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998357288079996,
      "loss": 5.7716,
      "mean_token_accuracy": 0.16418982148170472,
      "num_tokens": 5172100.0,
      "step": 2255
    },
    {
      "entropy": 5.9312032699584964,
      "epoch": 0.21709894332372717,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998344163798125,
      "loss": 5.7397,
      "mean_token_accuracy": 0.16633592247962953,
      "num_tokens": 5183984.0,
      "step": 2260
    },
    {
      "entropy": 5.862771272659302,
      "epoch": 0.21757925072046108,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998330987316379,
      "loss": 5.7188,
      "mean_token_accuracy": 0.1685562714934349,
      "num_tokens": 5195853.0,
      "step": 2265
    },
    {
      "entropy": 5.7872209548950195,
      "epoch": 0.21805955811719502,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998317758635062,
      "loss": 5.564,
      "mean_token_accuracy": 0.17655707597732545,
      "num_tokens": 5206995.0,
      "step": 2270
    },
    {
      "entropy": 5.662486600875854,
      "epoch": 0.21853986551392893,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998304477754484,
      "loss": 5.601,
      "mean_token_accuracy": 0.17818109691143036,
      "num_tokens": 5219291.0,
      "step": 2275
    },
    {
      "entropy": 5.872530078887939,
      "epoch": 0.21902017291066284,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998291144674952,
      "loss": 5.6762,
      "mean_token_accuracy": 0.17558231055736542,
      "num_tokens": 5230856.0,
      "step": 2280
    },
    {
      "entropy": 5.743492794036865,
      "epoch": 0.21950048030739674,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998277759396776,
      "loss": 5.5404,
      "mean_token_accuracy": 0.18575960099697114,
      "num_tokens": 5242871.0,
      "step": 2285
    },
    {
      "entropy": 5.753171300888061,
      "epoch": 0.21998078770413065,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998264321920265,
      "loss": 5.645,
      "mean_token_accuracy": 0.17795871943235397,
      "num_tokens": 5253835.0,
      "step": 2290
    },
    {
      "entropy": 5.834173107147217,
      "epoch": 0.22046109510086456,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998250832245734,
      "loss": 5.6285,
      "mean_token_accuracy": 0.17905001789331437,
      "num_tokens": 5266195.0,
      "step": 2295
    },
    {
      "entropy": 5.795336961746216,
      "epoch": 0.22094140249759847,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998237290373494,
      "loss": 5.6088,
      "mean_token_accuracy": 0.18106043636798858,
      "num_tokens": 5277499.0,
      "step": 2300
    },
    {
      "entropy": 5.926272821426392,
      "epoch": 0.22142170989433238,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499822369630386,
      "loss": 5.7316,
      "mean_token_accuracy": 0.16407669484615325,
      "num_tokens": 5288622.0,
      "step": 2305
    },
    {
      "entropy": 5.853435325622558,
      "epoch": 0.2219020172910663,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998210050037148,
      "loss": 5.7626,
      "mean_token_accuracy": 0.1640668675303459,
      "num_tokens": 5299664.0,
      "step": 2310
    },
    {
      "entropy": 5.848914766311646,
      "epoch": 0.2223823246878002,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998196351573674,
      "loss": 5.6597,
      "mean_token_accuracy": 0.17648861855268477,
      "num_tokens": 5311627.0,
      "step": 2315
    },
    {
      "entropy": 5.746734476089477,
      "epoch": 0.2228626320845341,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998182600913757,
      "loss": 5.5634,
      "mean_token_accuracy": 0.18326989710330963,
      "num_tokens": 5323000.0,
      "step": 2320
    },
    {
      "entropy": 5.806483554840088,
      "epoch": 0.22334293948126802,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998168798057715,
      "loss": 5.601,
      "mean_token_accuracy": 0.17859914302825927,
      "num_tokens": 5333811.0,
      "step": 2325
    },
    {
      "entropy": 5.761362361907959,
      "epoch": 0.22382324687800192,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499815494300587,
      "loss": 5.6037,
      "mean_token_accuracy": 0.17953155487775802,
      "num_tokens": 5344762.0,
      "step": 2330
    },
    {
      "entropy": 5.87151312828064,
      "epoch": 0.22430355427473583,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998141035758542,
      "loss": 5.6178,
      "mean_token_accuracy": 0.17849526405334473,
      "num_tokens": 5356112.0,
      "step": 2335
    },
    {
      "entropy": 5.816241693496704,
      "epoch": 0.22478386167146974,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004998127076316054,
      "loss": 5.7338,
      "mean_token_accuracy": 0.17089829891920089,
      "num_tokens": 5367339.0,
      "step": 2340
    },
    {
      "entropy": 5.803972244262695,
      "epoch": 0.22526416906820365,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998113064678734,
      "loss": 5.6492,
      "mean_token_accuracy": 0.1800484448671341,
      "num_tokens": 5378627.0,
      "step": 2345
    },
    {
      "entropy": 5.852473402023316,
      "epoch": 0.22574447646493756,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998099000846901,
      "loss": 5.6988,
      "mean_token_accuracy": 0.170039651542902,
      "num_tokens": 5390209.0,
      "step": 2350
    },
    {
      "entropy": 5.897865295410156,
      "epoch": 0.22622478386167147,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004998084884820887,
      "loss": 5.6853,
      "mean_token_accuracy": 0.17408402562141417,
      "num_tokens": 5401578.0,
      "step": 2355
    },
    {
      "entropy": 5.756231594085693,
      "epoch": 0.22670509125840538,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998070716601016,
      "loss": 5.604,
      "mean_token_accuracy": 0.17672595530748367,
      "num_tokens": 5413831.0,
      "step": 2360
    },
    {
      "entropy": 5.865112972259522,
      "epoch": 0.2271853986551393,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998056496187618,
      "loss": 5.6552,
      "mean_token_accuracy": 0.17331201881170272,
      "num_tokens": 5425430.0,
      "step": 2365
    },
    {
      "entropy": 5.671746349334716,
      "epoch": 0.2276657060518732,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998042223581025,
      "loss": 5.5077,
      "mean_token_accuracy": 0.18612135499715804,
      "num_tokens": 5435353.0,
      "step": 2370
    },
    {
      "entropy": 5.847486209869385,
      "epoch": 0.2281460134486071,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998027898781565,
      "loss": 5.7082,
      "mean_token_accuracy": 0.17188700735569,
      "num_tokens": 5446925.0,
      "step": 2375
    },
    {
      "entropy": 5.763326358795166,
      "epoch": 0.228626320845341,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998013521789574,
      "loss": 5.5835,
      "mean_token_accuracy": 0.18231132328510286,
      "num_tokens": 5456613.0,
      "step": 2380
    },
    {
      "entropy": 5.823438310623169,
      "epoch": 0.22910662824207492,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997999092605384,
      "loss": 5.6236,
      "mean_token_accuracy": 0.17775536924600602,
      "num_tokens": 5467790.0,
      "step": 2385
    },
    {
      "entropy": 5.813823795318603,
      "epoch": 0.22958693563880883,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499798461122933,
      "loss": 5.6084,
      "mean_token_accuracy": 0.17633272111415862,
      "num_tokens": 5479166.0,
      "step": 2390
    },
    {
      "entropy": 5.752776956558227,
      "epoch": 0.23006724303554274,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997970077661748,
      "loss": 5.592,
      "mean_token_accuracy": 0.18635759204626084,
      "num_tokens": 5490186.0,
      "step": 2395
    },
    {
      "entropy": 5.799315357208252,
      "epoch": 0.23054755043227665,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997955491902977,
      "loss": 5.581,
      "mean_token_accuracy": 0.173208612203598,
      "num_tokens": 5500416.0,
      "step": 2400
    },
    {
      "entropy": 5.78081750869751,
      "epoch": 0.23102785782901056,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997940853953354,
      "loss": 5.6455,
      "mean_token_accuracy": 0.17610323131084443,
      "num_tokens": 5512189.0,
      "step": 2405
    },
    {
      "entropy": 5.750943851470947,
      "epoch": 0.23150816522574447,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499792616381322,
      "loss": 5.5152,
      "mean_token_accuracy": 0.18103354275226594,
      "num_tokens": 5523631.0,
      "step": 2410
    },
    {
      "entropy": 5.769753503799438,
      "epoch": 0.23198847262247838,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997911421482914,
      "loss": 5.5793,
      "mean_token_accuracy": 0.18183256834745407,
      "num_tokens": 5535637.0,
      "step": 2415
    },
    {
      "entropy": 5.7926108837127686,
      "epoch": 0.23246878001921228,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499789662696278,
      "loss": 5.5809,
      "mean_token_accuracy": 0.18048014342784882,
      "num_tokens": 5546470.0,
      "step": 2420
    },
    {
      "entropy": 5.866393899917602,
      "epoch": 0.2329490874159462,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997881780253162,
      "loss": 5.7472,
      "mean_token_accuracy": 0.1731916069984436,
      "num_tokens": 5558633.0,
      "step": 2425
    },
    {
      "entropy": 5.785652208328247,
      "epoch": 0.2334293948126801,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004997866881354403,
      "loss": 5.6529,
      "mean_token_accuracy": 0.1741258069872856,
      "num_tokens": 5570427.0,
      "step": 2430
    },
    {
      "entropy": 5.814468336105347,
      "epoch": 0.23390970220941404,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499785193026685,
      "loss": 5.6383,
      "mean_token_accuracy": 0.17221823185682297,
      "num_tokens": 5580991.0,
      "step": 2435
    },
    {
      "entropy": 5.831542491912842,
      "epoch": 0.23439000960614795,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997836926990851,
      "loss": 5.688,
      "mean_token_accuracy": 0.1702160969376564,
      "num_tokens": 5592777.0,
      "step": 2440
    },
    {
      "entropy": 5.743597841262817,
      "epoch": 0.23487031700288186,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997821871526752,
      "loss": 5.5844,
      "mean_token_accuracy": 0.18186513036489488,
      "num_tokens": 5603326.0,
      "step": 2445
    },
    {
      "entropy": 5.7442912578582765,
      "epoch": 0.23535062439961577,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997806763874905,
      "loss": 5.5642,
      "mean_token_accuracy": 0.17944303005933762,
      "num_tokens": 5614504.0,
      "step": 2450
    },
    {
      "entropy": 5.771292018890381,
      "epoch": 0.23583093179634967,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997791604035659,
      "loss": 5.6286,
      "mean_token_accuracy": 0.18041289001703262,
      "num_tokens": 5625150.0,
      "step": 2455
    },
    {
      "entropy": 5.772236967086792,
      "epoch": 0.23631123919308358,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997776392009366,
      "loss": 5.6376,
      "mean_token_accuracy": 0.17271037846803666,
      "num_tokens": 5636815.0,
      "step": 2460
    },
    {
      "entropy": 5.824061107635498,
      "epoch": 0.2367915465898175,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997761127796381,
      "loss": 5.6228,
      "mean_token_accuracy": 0.17515492290258408,
      "num_tokens": 5648272.0,
      "step": 2465
    },
    {
      "entropy": 5.761957883834839,
      "epoch": 0.2372718539865514,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997745811397056,
      "loss": 5.5481,
      "mean_token_accuracy": 0.17955229729413985,
      "num_tokens": 5659227.0,
      "step": 2470
    },
    {
      "entropy": 5.7706011772155765,
      "epoch": 0.2377521613832853,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997730442811748,
      "loss": 5.668,
      "mean_token_accuracy": 0.1772313669323921,
      "num_tokens": 5670411.0,
      "step": 2475
    },
    {
      "entropy": 5.764390659332276,
      "epoch": 0.23823246878001922,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997715022040814,
      "loss": 5.5125,
      "mean_token_accuracy": 0.1818181738257408,
      "num_tokens": 5681570.0,
      "step": 2480
    },
    {
      "entropy": 5.693533134460449,
      "epoch": 0.23871277617675313,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499769954908461,
      "loss": 5.5075,
      "mean_token_accuracy": 0.18814098387956618,
      "num_tokens": 5693021.0,
      "step": 2485
    },
    {
      "entropy": 5.732751131057739,
      "epoch": 0.23919308357348704,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997684023943498,
      "loss": 5.5841,
      "mean_token_accuracy": 0.17531975656747817,
      "num_tokens": 5704043.0,
      "step": 2490
    },
    {
      "entropy": 5.747956085205078,
      "epoch": 0.23967339097022095,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997668446617837,
      "loss": 5.6703,
      "mean_token_accuracy": 0.17353179901838303,
      "num_tokens": 5715735.0,
      "step": 2495
    },
    {
      "entropy": 5.906291723251343,
      "epoch": 0.24015369836695485,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997652817107989,
      "loss": 5.6301,
      "mean_token_accuracy": 0.1773464038968086,
      "num_tokens": 5725778.0,
      "step": 2500
    },
    {
      "entropy": 5.7474383354187015,
      "epoch": 0.24063400576368876,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997637135414315,
      "loss": 5.6558,
      "mean_token_accuracy": 0.17596883922815323,
      "num_tokens": 5737224.0,
      "step": 2505
    },
    {
      "entropy": 5.883420705795288,
      "epoch": 0.24111431316042267,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997621401537183,
      "loss": 5.6777,
      "mean_token_accuracy": 0.17386398762464522,
      "num_tokens": 5749226.0,
      "step": 2510
    },
    {
      "entropy": 5.84810619354248,
      "epoch": 0.24159462055715658,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997605615476955,
      "loss": 5.6523,
      "mean_token_accuracy": 0.17698420137166976,
      "num_tokens": 5760282.0,
      "step": 2515
    },
    {
      "entropy": 5.712239980697632,
      "epoch": 0.2420749279538905,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997589777234,
      "loss": 5.5534,
      "mean_token_accuracy": 0.18517332524061203,
      "num_tokens": 5771756.0,
      "step": 2520
    },
    {
      "entropy": 5.766052198410034,
      "epoch": 0.2425552353506244,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997573886808684,
      "loss": 5.5824,
      "mean_token_accuracy": 0.1730706885457039,
      "num_tokens": 5783237.0,
      "step": 2525
    },
    {
      "entropy": 5.8274911403656,
      "epoch": 0.2430355427473583,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997557944201375,
      "loss": 5.6739,
      "mean_token_accuracy": 0.17468079626560212,
      "num_tokens": 5794825.0,
      "step": 2530
    },
    {
      "entropy": 5.793378686904907,
      "epoch": 0.24351585014409222,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997541949412445,
      "loss": 5.5834,
      "mean_token_accuracy": 0.18117998987436296,
      "num_tokens": 5805578.0,
      "step": 2535
    },
    {
      "entropy": 5.833753156661987,
      "epoch": 0.24399615754082613,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997525902442266,
      "loss": 5.6623,
      "mean_token_accuracy": 0.16987032443284988,
      "num_tokens": 5818201.0,
      "step": 2540
    },
    {
      "entropy": 5.754938840866089,
      "epoch": 0.24447646493756003,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997509803291207,
      "loss": 5.5809,
      "mean_token_accuracy": 0.17766104489564896,
      "num_tokens": 5830319.0,
      "step": 2545
    },
    {
      "entropy": 5.7188770294189455,
      "epoch": 0.24495677233429394,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997493651959647,
      "loss": 5.5319,
      "mean_token_accuracy": 0.18243667036294936,
      "num_tokens": 5840638.0,
      "step": 2550
    },
    {
      "entropy": 5.779051208496094,
      "epoch": 0.24543707973102785,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997477448447955,
      "loss": 5.5674,
      "mean_token_accuracy": 0.17579597234725952,
      "num_tokens": 5852472.0,
      "step": 2555
    },
    {
      "entropy": 5.779905652999878,
      "epoch": 0.24591738712776176,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997461192756512,
      "loss": 5.6071,
      "mean_token_accuracy": 0.17549769282341005,
      "num_tokens": 5863455.0,
      "step": 2560
    },
    {
      "entropy": 5.7139819145202635,
      "epoch": 0.24639769452449567,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997444884885694,
      "loss": 5.5082,
      "mean_token_accuracy": 0.1821136549115181,
      "num_tokens": 5873141.0,
      "step": 2565
    },
    {
      "entropy": 5.714353847503662,
      "epoch": 0.24687800192122958,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997428524835879,
      "loss": 5.6137,
      "mean_token_accuracy": 0.177882120013237,
      "num_tokens": 5884363.0,
      "step": 2570
    },
    {
      "entropy": 5.8017199516296385,
      "epoch": 0.2473583093179635,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997412112607446,
      "loss": 5.6563,
      "mean_token_accuracy": 0.17355215102434157,
      "num_tokens": 5895856.0,
      "step": 2575
    },
    {
      "entropy": 5.7232537269592285,
      "epoch": 0.2478386167146974,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997395648200778,
      "loss": 5.4999,
      "mean_token_accuracy": 0.18378839939832686,
      "num_tokens": 5906657.0,
      "step": 2580
    },
    {
      "entropy": 5.733435249328613,
      "epoch": 0.2483189241114313,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997379131616257,
      "loss": 5.6144,
      "mean_token_accuracy": 0.17644155025482178,
      "num_tokens": 5919496.0,
      "step": 2585
    },
    {
      "entropy": 5.843182802200317,
      "epoch": 0.24879923150816521,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997362562854266,
      "loss": 5.6782,
      "mean_token_accuracy": 0.16554286181926728,
      "num_tokens": 5932593.0,
      "step": 2590
    },
    {
      "entropy": 5.776597499847412,
      "epoch": 0.24927953890489912,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997345941915187,
      "loss": 5.5987,
      "mean_token_accuracy": 0.1750054731965065,
      "num_tokens": 5944080.0,
      "step": 2595
    },
    {
      "entropy": 5.734813213348389,
      "epoch": 0.24975984630163303,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997329268799412,
      "loss": 5.5622,
      "mean_token_accuracy": 0.18719092905521392,
      "num_tokens": 5955703.0,
      "step": 2600
    },
    {
      "entropy": 5.736793804168701,
      "epoch": 0.25024015369836694,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997312543507322,
      "loss": 5.6578,
      "mean_token_accuracy": 0.1708198830485344,
      "num_tokens": 5966979.0,
      "step": 2605
    },
    {
      "entropy": 5.811160707473755,
      "epoch": 0.2507204610951009,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997295766039309,
      "loss": 5.5433,
      "mean_token_accuracy": 0.17897313684225083,
      "num_tokens": 5978808.0,
      "step": 2610
    },
    {
      "entropy": 5.785584354400635,
      "epoch": 0.25120076849183476,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997278936395761,
      "loss": 5.7051,
      "mean_token_accuracy": 0.16843380331993102,
      "num_tokens": 5992145.0,
      "step": 2615
    },
    {
      "entropy": 5.7584997653961185,
      "epoch": 0.2516810758885687,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997262054577071,
      "loss": 5.5743,
      "mean_token_accuracy": 0.17471926659345627,
      "num_tokens": 6003723.0,
      "step": 2620
    },
    {
      "entropy": 5.771317625045777,
      "epoch": 0.2521613832853026,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997245120583627,
      "loss": 5.6058,
      "mean_token_accuracy": 0.18242801427841188,
      "num_tokens": 6014064.0,
      "step": 2625
    },
    {
      "entropy": 5.698479509353637,
      "epoch": 0.2526416906820365,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997228134415825,
      "loss": 5.5187,
      "mean_token_accuracy": 0.18279289156198503,
      "num_tokens": 6025455.0,
      "step": 2630
    },
    {
      "entropy": 5.782511425018311,
      "epoch": 0.2531219980787704,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997211096074059,
      "loss": 5.6313,
      "mean_token_accuracy": 0.1700494483113289,
      "num_tokens": 6037347.0,
      "step": 2635
    },
    {
      "entropy": 5.6879799365997314,
      "epoch": 0.25360230547550433,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997194005558722,
      "loss": 5.5155,
      "mean_token_accuracy": 0.18077107816934584,
      "num_tokens": 6049236.0,
      "step": 2640
    },
    {
      "entropy": 5.664460229873657,
      "epoch": 0.2540826128722382,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997176862870216,
      "loss": 5.5396,
      "mean_token_accuracy": 0.1818702682852745,
      "num_tokens": 6060982.0,
      "step": 2645
    },
    {
      "entropy": 5.795065593719483,
      "epoch": 0.25456292026897215,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997159668008933,
      "loss": 5.5409,
      "mean_token_accuracy": 0.18014705032110215,
      "num_tokens": 6070925.0,
      "step": 2650
    },
    {
      "entropy": 5.672208833694458,
      "epoch": 0.25504322766570603,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997142420975277,
      "loss": 5.505,
      "mean_token_accuracy": 0.18191429376602172,
      "num_tokens": 6081279.0,
      "step": 2655
    },
    {
      "entropy": 5.707340812683105,
      "epoch": 0.25552353506243997,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997125121769647,
      "loss": 5.5966,
      "mean_token_accuracy": 0.17941437065601348,
      "num_tokens": 6091797.0,
      "step": 2660
    },
    {
      "entropy": 5.816579580307007,
      "epoch": 0.25600384245917385,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997107770392444,
      "loss": 5.6217,
      "mean_token_accuracy": 0.18266446739435196,
      "num_tokens": 6103435.0,
      "step": 2665
    },
    {
      "entropy": 5.8208917617797855,
      "epoch": 0.2564841498559078,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499709036684407,
      "loss": 5.6735,
      "mean_token_accuracy": 0.1784491568803787,
      "num_tokens": 6114531.0,
      "step": 2670
    },
    {
      "entropy": 5.659421300888061,
      "epoch": 0.25696445725264166,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997072911124932,
      "loss": 5.5132,
      "mean_token_accuracy": 0.1838804692029953,
      "num_tokens": 6126110.0,
      "step": 2675
    },
    {
      "entropy": 5.79349536895752,
      "epoch": 0.2574447646493756,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997055403235432,
      "loss": 5.6132,
      "mean_token_accuracy": 0.17701641321182252,
      "num_tokens": 6137114.0,
      "step": 2680
    },
    {
      "entropy": 5.69478325843811,
      "epoch": 0.2579250720461095,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997037843175978,
      "loss": 5.5683,
      "mean_token_accuracy": 0.17944201529026033,
      "num_tokens": 6148696.0,
      "step": 2685
    },
    {
      "entropy": 5.735519981384277,
      "epoch": 0.2584053794428434,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997020230946978,
      "loss": 5.5677,
      "mean_token_accuracy": 0.18062195777893067,
      "num_tokens": 6160235.0,
      "step": 2690
    },
    {
      "entropy": 5.747908687591552,
      "epoch": 0.25888568683957736,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997002566548841,
      "loss": 5.5407,
      "mean_token_accuracy": 0.17935636788606643,
      "num_tokens": 6172031.0,
      "step": 2695
    },
    {
      "entropy": 5.596827173233033,
      "epoch": 0.25936599423631124,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996984849981976,
      "loss": 5.4283,
      "mean_token_accuracy": 0.18828413337469102,
      "num_tokens": 6183547.0,
      "step": 2700
    },
    {
      "entropy": 5.798890924453735,
      "epoch": 0.2598463016330452,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996967081246794,
      "loss": 5.6161,
      "mean_token_accuracy": 0.17421343475580214,
      "num_tokens": 6194768.0,
      "step": 2705
    },
    {
      "entropy": 5.752058315277099,
      "epoch": 0.26032660902977905,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996949260343711,
      "loss": 5.6108,
      "mean_token_accuracy": 0.17473251819610597,
      "num_tokens": 6206099.0,
      "step": 2710
    },
    {
      "entropy": 5.768714952468872,
      "epoch": 0.260806916426513,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996931387273137,
      "loss": 5.6049,
      "mean_token_accuracy": 0.18249890506267546,
      "num_tokens": 6217530.0,
      "step": 2715
    },
    {
      "entropy": 5.824987697601318,
      "epoch": 0.2612872238232469,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996913462035487,
      "loss": 5.6404,
      "mean_token_accuracy": 0.1773010641336441,
      "num_tokens": 6228564.0,
      "step": 2720
    },
    {
      "entropy": 5.682008457183838,
      "epoch": 0.2617675312199808,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499689548463118,
      "loss": 5.5072,
      "mean_token_accuracy": 0.18143168538808824,
      "num_tokens": 6239945.0,
      "step": 2725
    },
    {
      "entropy": 5.719662475585937,
      "epoch": 0.2622478386167147,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996877455060631,
      "loss": 5.6148,
      "mean_token_accuracy": 0.17328692674636842,
      "num_tokens": 6251829.0,
      "step": 2730
    },
    {
      "entropy": 5.861980676651001,
      "epoch": 0.2627281460134486,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996859373324259,
      "loss": 5.7102,
      "mean_token_accuracy": 0.1663394570350647,
      "num_tokens": 6264823.0,
      "step": 2735
    },
    {
      "entropy": 5.684702014923095,
      "epoch": 0.2632084534101825,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996841239422485,
      "loss": 5.4084,
      "mean_token_accuracy": 0.18651311248540878,
      "num_tokens": 6276247.0,
      "step": 2740
    },
    {
      "entropy": 5.626027154922485,
      "epoch": 0.26368876080691644,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996823053355729,
      "loss": 5.5308,
      "mean_token_accuracy": 0.18354606330394746,
      "num_tokens": 6287593.0,
      "step": 2745
    },
    {
      "entropy": 5.799522447586059,
      "epoch": 0.2641690682036503,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996804815124413,
      "loss": 5.676,
      "mean_token_accuracy": 0.17569343596696854,
      "num_tokens": 6299918.0,
      "step": 2750
    },
    {
      "entropy": 5.681373929977417,
      "epoch": 0.26464937560038426,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996786524728962,
      "loss": 5.5314,
      "mean_token_accuracy": 0.18340865522623062,
      "num_tokens": 6311147.0,
      "step": 2755
    },
    {
      "entropy": 5.615675544738769,
      "epoch": 0.26512968299711814,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996768182169797,
      "loss": 5.4432,
      "mean_token_accuracy": 0.18849314153194427,
      "num_tokens": 6323239.0,
      "step": 2760
    },
    {
      "entropy": 5.773361492156982,
      "epoch": 0.2656099903938521,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996749787447349,
      "loss": 5.5348,
      "mean_token_accuracy": 0.1791366621851921,
      "num_tokens": 6334625.0,
      "step": 2765
    },
    {
      "entropy": 5.70220251083374,
      "epoch": 0.26609029779058596,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499673134056204,
      "loss": 5.5651,
      "mean_token_accuracy": 0.17963075488805771,
      "num_tokens": 6346068.0,
      "step": 2770
    },
    {
      "entropy": 5.713240051269532,
      "epoch": 0.2665706051873199,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996712841514303,
      "loss": 5.5531,
      "mean_token_accuracy": 0.18010566383600235,
      "num_tokens": 6357097.0,
      "step": 2775
    },
    {
      "entropy": 5.7573648452758786,
      "epoch": 0.2670509125840538,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996694290304563,
      "loss": 5.6084,
      "mean_token_accuracy": 0.17204165309667588,
      "num_tokens": 6367481.0,
      "step": 2780
    },
    {
      "entropy": 5.685487174987793,
      "epoch": 0.2675312199807877,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996675686933255,
      "loss": 5.5207,
      "mean_token_accuracy": 0.18194421827793122,
      "num_tokens": 6378873.0,
      "step": 2785
    },
    {
      "entropy": 5.747344875335694,
      "epoch": 0.2680115273775216,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996657031400807,
      "loss": 5.5625,
      "mean_token_accuracy": 0.1808866173028946,
      "num_tokens": 6390651.0,
      "step": 2790
    },
    {
      "entropy": 5.640884160995483,
      "epoch": 0.26849183477425553,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996638323707655,
      "loss": 5.4351,
      "mean_token_accuracy": 0.18793859630823134,
      "num_tokens": 6401631.0,
      "step": 2795
    },
    {
      "entropy": 5.6249631404876705,
      "epoch": 0.2689721421709894,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996619563854232,
      "loss": 5.5219,
      "mean_token_accuracy": 0.186911478638649,
      "num_tokens": 6413875.0,
      "step": 2800
    },
    {
      "entropy": 5.782164573669434,
      "epoch": 0.26945244956772335,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996600751840974,
      "loss": 5.5488,
      "mean_token_accuracy": 0.17567767947912216,
      "num_tokens": 6425764.0,
      "step": 2805
    },
    {
      "entropy": 5.648550701141358,
      "epoch": 0.26993275696445723,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996581887668317,
      "loss": 5.4617,
      "mean_token_accuracy": 0.1833810329437256,
      "num_tokens": 6437911.0,
      "step": 2810
    },
    {
      "entropy": 5.6579090595245365,
      "epoch": 0.27041306436119117,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049965629713367,
      "loss": 5.4874,
      "mean_token_accuracy": 0.1852079749107361,
      "num_tokens": 6449942.0,
      "step": 2815
    },
    {
      "entropy": 5.709984302520752,
      "epoch": 0.27089337175792505,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996544002846561,
      "loss": 5.6016,
      "mean_token_accuracy": 0.18187092244625092,
      "num_tokens": 6461729.0,
      "step": 2820
    },
    {
      "entropy": 5.747480726242065,
      "epoch": 0.271373679154659,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996524982198343,
      "loss": 5.5839,
      "mean_token_accuracy": 0.18295794278383254,
      "num_tokens": 6472046.0,
      "step": 2825
    },
    {
      "entropy": 5.714561557769775,
      "epoch": 0.27185398655139287,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996505909392485,
      "loss": 5.5436,
      "mean_token_accuracy": 0.1794055625796318,
      "num_tokens": 6483308.0,
      "step": 2830
    },
    {
      "entropy": 5.5759443759918215,
      "epoch": 0.2723342939481268,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996486784429429,
      "loss": 5.4213,
      "mean_token_accuracy": 0.18884292393922805,
      "num_tokens": 6495093.0,
      "step": 2835
    },
    {
      "entropy": 5.70222053527832,
      "epoch": 0.2728146013448607,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996467607309622,
      "loss": 5.5213,
      "mean_token_accuracy": 0.18563241511583328,
      "num_tokens": 6505933.0,
      "step": 2840
    },
    {
      "entropy": 5.663483715057373,
      "epoch": 0.2732949087415946,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996448378033507,
      "loss": 5.56,
      "mean_token_accuracy": 0.17418374568223954,
      "num_tokens": 6517280.0,
      "step": 2845
    },
    {
      "entropy": 5.739472961425781,
      "epoch": 0.2737752161383285,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996429096601532,
      "loss": 5.62,
      "mean_token_accuracy": 0.1724501520395279,
      "num_tokens": 6528980.0,
      "step": 2850
    },
    {
      "entropy": 5.694519329071045,
      "epoch": 0.27425552353506244,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996409763014144,
      "loss": 5.6063,
      "mean_token_accuracy": 0.18012285381555557,
      "num_tokens": 6540670.0,
      "step": 2855
    },
    {
      "entropy": 5.6691272258758545,
      "epoch": 0.2747358309317964,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996390377271791,
      "loss": 5.5641,
      "mean_token_accuracy": 0.18185704350471496,
      "num_tokens": 6551302.0,
      "step": 2860
    },
    {
      "entropy": 5.717091369628906,
      "epoch": 0.27521613832853026,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996370939374924,
      "loss": 5.5206,
      "mean_token_accuracy": 0.17356996536254882,
      "num_tokens": 6563177.0,
      "step": 2865
    },
    {
      "entropy": 5.848795604705811,
      "epoch": 0.2756964457252642,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996351449323994,
      "loss": 5.6331,
      "mean_token_accuracy": 0.18121621906757354,
      "num_tokens": 6573323.0,
      "step": 2870
    },
    {
      "entropy": 5.6664739608764645,
      "epoch": 0.2761767531219981,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996331907119455,
      "loss": 5.5584,
      "mean_token_accuracy": 0.17100237160921097,
      "num_tokens": 6585382.0,
      "step": 2875
    },
    {
      "entropy": 5.612715768814087,
      "epoch": 0.276657060518732,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996312312761758,
      "loss": 5.4473,
      "mean_token_accuracy": 0.18974230736494063,
      "num_tokens": 6596629.0,
      "step": 2880
    },
    {
      "entropy": 5.721270322799683,
      "epoch": 0.2771373679154659,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499629266625136,
      "loss": 5.5591,
      "mean_token_accuracy": 0.17991002202033995,
      "num_tokens": 6608408.0,
      "step": 2885
    },
    {
      "entropy": 5.779607629776001,
      "epoch": 0.27761767531219983,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996272967588715,
      "loss": 5.5857,
      "mean_token_accuracy": 0.17561327517032624,
      "num_tokens": 6619375.0,
      "step": 2890
    },
    {
      "entropy": 5.758464813232422,
      "epoch": 0.2780979827089337,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996253216774283,
      "loss": 5.6217,
      "mean_token_accuracy": 0.17862538397312164,
      "num_tokens": 6631317.0,
      "step": 2895
    },
    {
      "entropy": 5.706511449813843,
      "epoch": 0.27857829010566765,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996233413808521,
      "loss": 5.4881,
      "mean_token_accuracy": 0.18149750381708146,
      "num_tokens": 6642009.0,
      "step": 2900
    },
    {
      "entropy": 5.686898136138916,
      "epoch": 0.27905859750240153,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996213558691889,
      "loss": 5.6358,
      "mean_token_accuracy": 0.17196063697338104,
      "num_tokens": 6654713.0,
      "step": 2905
    },
    {
      "entropy": 5.74081768989563,
      "epoch": 0.27953890489913547,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996193651424848,
      "loss": 5.5878,
      "mean_token_accuracy": 0.1804736316204071,
      "num_tokens": 6667157.0,
      "step": 2910
    },
    {
      "entropy": 5.703048181533814,
      "epoch": 0.28001921229586935,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499617369200786,
      "loss": 5.5329,
      "mean_token_accuracy": 0.1874094307422638,
      "num_tokens": 6679573.0,
      "step": 2915
    },
    {
      "entropy": 5.73606538772583,
      "epoch": 0.2804995196926033,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004996153680441389,
      "loss": 5.6154,
      "mean_token_accuracy": 0.1785501942038536,
      "num_tokens": 6691768.0,
      "step": 2920
    },
    {
      "entropy": 5.75312385559082,
      "epoch": 0.28097982708933716,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049961336167259,
      "loss": 5.5727,
      "mean_token_accuracy": 0.1757212147116661,
      "num_tokens": 6701964.0,
      "step": 2925
    },
    {
      "entropy": 5.748111152648926,
      "epoch": 0.2814601344860711,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996113500861857,
      "loss": 5.5561,
      "mean_token_accuracy": 0.18022931665182113,
      "num_tokens": 6713506.0,
      "step": 2930
    },
    {
      "entropy": 5.749008369445801,
      "epoch": 0.281940441882805,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996093332849729,
      "loss": 5.5821,
      "mean_token_accuracy": 0.17544898986816407,
      "num_tokens": 6724616.0,
      "step": 2935
    },
    {
      "entropy": 5.6963791847229,
      "epoch": 0.2824207492795389,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996073112689983,
      "loss": 5.5629,
      "mean_token_accuracy": 0.18043071925640106,
      "num_tokens": 6735054.0,
      "step": 2940
    },
    {
      "entropy": 5.741255426406861,
      "epoch": 0.2829010566762728,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996052840383088,
      "loss": 5.6018,
      "mean_token_accuracy": 0.1793173149228096,
      "num_tokens": 6746756.0,
      "step": 2945
    },
    {
      "entropy": 5.704806613922119,
      "epoch": 0.28338136407300674,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004996032515929516,
      "loss": 5.4832,
      "mean_token_accuracy": 0.1823109433054924,
      "num_tokens": 6759566.0,
      "step": 2950
    },
    {
      "entropy": 5.69014163017273,
      "epoch": 0.2838616714697406,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996012139329738,
      "loss": 5.5125,
      "mean_token_accuracy": 0.18087273538112641,
      "num_tokens": 6771375.0,
      "step": 2955
    },
    {
      "entropy": 5.748940420150757,
      "epoch": 0.28434197886647455,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995991710584228,
      "loss": 5.6176,
      "mean_token_accuracy": 0.1698983356356621,
      "num_tokens": 6783252.0,
      "step": 2960
    },
    {
      "entropy": 5.6928369998931885,
      "epoch": 0.28482228626320844,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995971229693459,
      "loss": 5.5816,
      "mean_token_accuracy": 0.1742567628622055,
      "num_tokens": 6795525.0,
      "step": 2965
    },
    {
      "entropy": 5.742653036117554,
      "epoch": 0.28530259365994237,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995950696657909,
      "loss": 5.5129,
      "mean_token_accuracy": 0.18347676545381547,
      "num_tokens": 6807212.0,
      "step": 2970
    },
    {
      "entropy": 5.64953875541687,
      "epoch": 0.28578290105667625,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995930111478051,
      "loss": 5.458,
      "mean_token_accuracy": 0.18245918452739715,
      "num_tokens": 6819367.0,
      "step": 2975
    },
    {
      "entropy": 5.665671300888062,
      "epoch": 0.2862632084534102,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995909474154365,
      "loss": 5.5454,
      "mean_token_accuracy": 0.18168749809265136,
      "num_tokens": 6830405.0,
      "step": 2980
    },
    {
      "entropy": 5.6374242305755615,
      "epoch": 0.28674351585014407,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995888784687331,
      "loss": 5.513,
      "mean_token_accuracy": 0.1782350167632103,
      "num_tokens": 6841479.0,
      "step": 2985
    },
    {
      "entropy": 5.671642160415649,
      "epoch": 0.287223823246878,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995868043077428,
      "loss": 5.5612,
      "mean_token_accuracy": 0.17904469966888428,
      "num_tokens": 6851585.0,
      "step": 2990
    },
    {
      "entropy": 5.694643163681031,
      "epoch": 0.2877041306436119,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995847249325137,
      "loss": 5.521,
      "mean_token_accuracy": 0.18657984137535094,
      "num_tokens": 6863176.0,
      "step": 2995
    },
    {
      "entropy": 5.726582813262939,
      "epoch": 0.2881844380403458,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995826403430942,
      "loss": 5.5769,
      "mean_token_accuracy": 0.1821230873465538,
      "num_tokens": 6874021.0,
      "step": 3000
    },
    {
      "epoch": 0.2881844380403458,
      "eval_entropy": 5.499006296238743,
      "eval_loss": 5.545695781707764,
      "eval_mean_token_accuracy": 0.1887321266854706,
      "eval_num_tokens": 6874021.0,
      "eval_runtime": 20.7487,
      "eval_samples_per_second": 1581.543,
      "eval_steps_per_second": 197.699,
      "step": 3000
    },
    {
      "entropy": 5.730997228622437,
      "epoch": 0.2886647454370797,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995805505395328,
      "loss": 5.5433,
      "mean_token_accuracy": 0.17902631163597107,
      "num_tokens": 6884999.0,
      "step": 3005
    },
    {
      "entropy": 5.664561748504639,
      "epoch": 0.28914505283381364,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995784555218778,
      "loss": 5.5182,
      "mean_token_accuracy": 0.1828203856945038,
      "num_tokens": 6897021.0,
      "step": 3010
    },
    {
      "entropy": 5.648147296905518,
      "epoch": 0.2896253602305475,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995763552901779,
      "loss": 5.5212,
      "mean_token_accuracy": 0.18067535609006882,
      "num_tokens": 6908320.0,
      "step": 3015
    },
    {
      "entropy": 5.756967782974243,
      "epoch": 0.29010566762728146,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995742498444818,
      "loss": 5.5074,
      "mean_token_accuracy": 0.1846700832247734,
      "num_tokens": 6919957.0,
      "step": 3020
    },
    {
      "entropy": 5.698904705047608,
      "epoch": 0.2905859750240154,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995721391848387,
      "loss": 5.4824,
      "mean_token_accuracy": 0.17627989798784255,
      "num_tokens": 6930531.0,
      "step": 3025
    },
    {
      "entropy": 5.720267581939697,
      "epoch": 0.2910662824207493,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995700233112972,
      "loss": 5.6076,
      "mean_token_accuracy": 0.17746972441673278,
      "num_tokens": 6942556.0,
      "step": 3030
    },
    {
      "entropy": 5.70178074836731,
      "epoch": 0.2915465898174832,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995679022239066,
      "loss": 5.5621,
      "mean_token_accuracy": 0.18203302770853041,
      "num_tokens": 6954410.0,
      "step": 3035
    },
    {
      "entropy": 5.717909717559815,
      "epoch": 0.2920268972142171,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995657759227162,
      "loss": 5.5576,
      "mean_token_accuracy": 0.1767730161547661,
      "num_tokens": 6964970.0,
      "step": 3040
    },
    {
      "entropy": 5.675133419036865,
      "epoch": 0.29250720461095103,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995636444077751,
      "loss": 5.4469,
      "mean_token_accuracy": 0.18839906603097917,
      "num_tokens": 6976016.0,
      "step": 3045
    },
    {
      "entropy": 5.634436845779419,
      "epoch": 0.2929875120076849,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995615076791333,
      "loss": 5.4858,
      "mean_token_accuracy": 0.18708995282649993,
      "num_tokens": 6987199.0,
      "step": 3050
    },
    {
      "entropy": 5.65491418838501,
      "epoch": 0.29346781940441885,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995593657368399,
      "loss": 5.4941,
      "mean_token_accuracy": 0.19084091037511824,
      "num_tokens": 6999174.0,
      "step": 3055
    },
    {
      "entropy": 5.698542308807373,
      "epoch": 0.29394812680115273,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499557218580945,
      "loss": 5.5702,
      "mean_token_accuracy": 0.18034818172454833,
      "num_tokens": 7012148.0,
      "step": 3060
    },
    {
      "entropy": 5.570310926437378,
      "epoch": 0.29442843419788667,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995550662114981,
      "loss": 5.4448,
      "mean_token_accuracy": 0.18498128652572632,
      "num_tokens": 7023238.0,
      "step": 3065
    },
    {
      "entropy": 5.627264976501465,
      "epoch": 0.29490874159462055,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995529086285495,
      "loss": 5.5159,
      "mean_token_accuracy": 0.1867678791284561,
      "num_tokens": 7034944.0,
      "step": 3070
    },
    {
      "entropy": 5.691465091705322,
      "epoch": 0.2953890489913545,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499550745832149,
      "loss": 5.3885,
      "mean_token_accuracy": 0.19256407022476196,
      "num_tokens": 7046880.0,
      "step": 3075
    },
    {
      "entropy": 5.630908727645874,
      "epoch": 0.29586935638808837,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995485778223471,
      "loss": 5.4752,
      "mean_token_accuracy": 0.18294412046670913,
      "num_tokens": 7057678.0,
      "step": 3080
    },
    {
      "entropy": 5.569836091995239,
      "epoch": 0.2963496637848223,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995464045991939,
      "loss": 5.443,
      "mean_token_accuracy": 0.1914595991373062,
      "num_tokens": 7068336.0,
      "step": 3085
    },
    {
      "entropy": 5.677788400650025,
      "epoch": 0.2968299711815562,
      "grad_norm": 0.890625,
      "learning_rate": 0.00049954422616274,
      "loss": 5.5156,
      "mean_token_accuracy": 0.17752356827259064,
      "num_tokens": 7080341.0,
      "step": 3090
    },
    {
      "entropy": 5.829857444763183,
      "epoch": 0.2973102785782901,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995420425130359,
      "loss": 5.6642,
      "mean_token_accuracy": 0.17347445338964462,
      "num_tokens": 7090618.0,
      "step": 3095
    },
    {
      "entropy": 5.578703689575195,
      "epoch": 0.297790585975024,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995398536501324,
      "loss": 5.4171,
      "mean_token_accuracy": 0.19119216352701188,
      "num_tokens": 7101843.0,
      "step": 3100
    },
    {
      "entropy": 5.6363883972167965,
      "epoch": 0.29827089337175794,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995376595740801,
      "loss": 5.4874,
      "mean_token_accuracy": 0.1814226746559143,
      "num_tokens": 7112014.0,
      "step": 3105
    },
    {
      "entropy": 5.710436534881592,
      "epoch": 0.2987512007684918,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995354602849302,
      "loss": 5.5624,
      "mean_token_accuracy": 0.17562751471996307,
      "num_tokens": 7123860.0,
      "step": 3110
    },
    {
      "entropy": 5.6648670673370365,
      "epoch": 0.29923150816522576,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995332557827337,
      "loss": 5.5357,
      "mean_token_accuracy": 0.18224934190511705,
      "num_tokens": 7135901.0,
      "step": 3115
    },
    {
      "entropy": 5.694220733642578,
      "epoch": 0.29971181556195964,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995310460675416,
      "loss": 5.4932,
      "mean_token_accuracy": 0.19112819135189058,
      "num_tokens": 7148743.0,
      "step": 3120
    },
    {
      "entropy": 5.702276515960693,
      "epoch": 0.3001921229586936,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995288311394053,
      "loss": 5.5516,
      "mean_token_accuracy": 0.18374135941267014,
      "num_tokens": 7160731.0,
      "step": 3125
    },
    {
      "entropy": 5.703460311889648,
      "epoch": 0.30067243035542746,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995266109983764,
      "loss": 5.5461,
      "mean_token_accuracy": 0.181298191845417,
      "num_tokens": 7172861.0,
      "step": 3130
    },
    {
      "entropy": 5.636017751693726,
      "epoch": 0.3011527377521614,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995243856445062,
      "loss": 5.4796,
      "mean_token_accuracy": 0.18060984313488007,
      "num_tokens": 7183954.0,
      "step": 3135
    },
    {
      "entropy": 5.591547632217408,
      "epoch": 0.3016330451488953,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995221550778466,
      "loss": 5.4632,
      "mean_token_accuracy": 0.18435155898332595,
      "num_tokens": 7195466.0,
      "step": 3140
    },
    {
      "entropy": 5.671104192733765,
      "epoch": 0.3021133525456292,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995199192984491,
      "loss": 5.45,
      "mean_token_accuracy": 0.18866306990385057,
      "num_tokens": 7207173.0,
      "step": 3145
    },
    {
      "entropy": 5.707768821716309,
      "epoch": 0.3025936599423631,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995176783063657,
      "loss": 5.5867,
      "mean_token_accuracy": 0.18185509890317916,
      "num_tokens": 7220095.0,
      "step": 3150
    },
    {
      "entropy": 5.608586359024048,
      "epoch": 0.30307396733909703,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995154321016487,
      "loss": 5.5059,
      "mean_token_accuracy": 0.18203232288360596,
      "num_tokens": 7230664.0,
      "step": 3155
    },
    {
      "entropy": 5.678262376785279,
      "epoch": 0.3035542747358309,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995131806843499,
      "loss": 5.4528,
      "mean_token_accuracy": 0.19190701693296433,
      "num_tokens": 7241278.0,
      "step": 3160
    },
    {
      "entropy": 5.556720972061157,
      "epoch": 0.30403458213256485,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995109240545218,
      "loss": 5.6049,
      "mean_token_accuracy": 0.17547910213470458,
      "num_tokens": 7252999.0,
      "step": 3165
    },
    {
      "entropy": 5.738193988800049,
      "epoch": 0.3045148895292987,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995086622122167,
      "loss": 5.4516,
      "mean_token_accuracy": 0.18523052483797073,
      "num_tokens": 7263949.0,
      "step": 3170
    },
    {
      "entropy": 5.610001945495606,
      "epoch": 0.30499519692603266,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995063951574871,
      "loss": 5.4864,
      "mean_token_accuracy": 0.18810611367225646,
      "num_tokens": 7275467.0,
      "step": 3175
    },
    {
      "entropy": 5.6769002914428714,
      "epoch": 0.30547550432276654,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995041228903856,
      "loss": 5.4697,
      "mean_token_accuracy": 0.1874818757176399,
      "num_tokens": 7285534.0,
      "step": 3180
    },
    {
      "entropy": 5.706569576263428,
      "epoch": 0.3059558117195005,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499501845410965,
      "loss": 5.5873,
      "mean_token_accuracy": 0.18355770856142045,
      "num_tokens": 7297252.0,
      "step": 3185
    },
    {
      "entropy": 5.64103512763977,
      "epoch": 0.30643611911623436,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994995627192781,
      "loss": 5.443,
      "mean_token_accuracy": 0.1906636580824852,
      "num_tokens": 7308492.0,
      "step": 3190
    },
    {
      "entropy": 5.620066118240357,
      "epoch": 0.3069164265129683,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994972748153781,
      "loss": 5.4835,
      "mean_token_accuracy": 0.1803540736436844,
      "num_tokens": 7319703.0,
      "step": 3195
    },
    {
      "entropy": 5.6844628810882565,
      "epoch": 0.30739673390970224,
      "grad_norm": 1.0,
      "learning_rate": 0.000499494981699318,
      "loss": 5.4619,
      "mean_token_accuracy": 0.18943321704864502,
      "num_tokens": 7331022.0,
      "step": 3200
    },
    {
      "entropy": 5.6341009616851805,
      "epoch": 0.3078770413064361,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499492683371151,
      "loss": 5.505,
      "mean_token_accuracy": 0.18882815837860106,
      "num_tokens": 7342977.0,
      "step": 3205
    },
    {
      "entropy": 5.703154134750366,
      "epoch": 0.30835734870317005,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994903798309306,
      "loss": 5.4818,
      "mean_token_accuracy": 0.18102553188800813,
      "num_tokens": 7353227.0,
      "step": 3210
    },
    {
      "entropy": 5.651052856445313,
      "epoch": 0.30883765609990393,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994880710787102,
      "loss": 5.5372,
      "mean_token_accuracy": 0.16935457289218903,
      "num_tokens": 7364165.0,
      "step": 3215
    },
    {
      "entropy": 5.601374197006225,
      "epoch": 0.30931796349663787,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994857571145432,
      "loss": 5.4713,
      "mean_token_accuracy": 0.18606537580490112,
      "num_tokens": 7374800.0,
      "step": 3220
    },
    {
      "entropy": 5.573496866226196,
      "epoch": 0.30979827089337175,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994834379384837,
      "loss": 5.4187,
      "mean_token_accuracy": 0.18962224274873735,
      "num_tokens": 7386360.0,
      "step": 3225
    },
    {
      "entropy": 5.631180143356323,
      "epoch": 0.3102785782901057,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994811135505851,
      "loss": 5.4459,
      "mean_token_accuracy": 0.1859535649418831,
      "num_tokens": 7397066.0,
      "step": 3230
    },
    {
      "entropy": 5.6822919845581055,
      "epoch": 0.31075888568683957,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994787839509018,
      "loss": 5.5174,
      "mean_token_accuracy": 0.17699058800935746,
      "num_tokens": 7408349.0,
      "step": 3235
    },
    {
      "entropy": 5.69844274520874,
      "epoch": 0.3112391930835735,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994764491394876,
      "loss": 5.5627,
      "mean_token_accuracy": 0.17459277361631392,
      "num_tokens": 7420343.0,
      "step": 3240
    },
    {
      "entropy": 5.702674007415771,
      "epoch": 0.3117195004803074,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004994741091163969,
      "loss": 5.4782,
      "mean_token_accuracy": 0.1868354335427284,
      "num_tokens": 7431683.0,
      "step": 3245
    },
    {
      "entropy": 5.497278642654419,
      "epoch": 0.3121998078770413,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499471763881684,
      "loss": 5.3854,
      "mean_token_accuracy": 0.18908552974462509,
      "num_tokens": 7443327.0,
      "step": 3250
    },
    {
      "entropy": 5.6632318019866945,
      "epoch": 0.3126801152737752,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994694134354031,
      "loss": 5.4845,
      "mean_token_accuracy": 0.18859678357839585,
      "num_tokens": 7454002.0,
      "step": 3255
    },
    {
      "entropy": 5.57216854095459,
      "epoch": 0.31316042267050914,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499467057777609,
      "loss": 5.4814,
      "mean_token_accuracy": 0.1825019970536232,
      "num_tokens": 7464074.0,
      "step": 3260
    },
    {
      "entropy": 5.594534063339234,
      "epoch": 0.313640730067243,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994646969083565,
      "loss": 5.4098,
      "mean_token_accuracy": 0.19650202989578247,
      "num_tokens": 7475543.0,
      "step": 3265
    },
    {
      "entropy": 5.698621129989624,
      "epoch": 0.31412103746397696,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994623308277002,
      "loss": 5.468,
      "mean_token_accuracy": 0.1819482773542404,
      "num_tokens": 7486818.0,
      "step": 3270
    },
    {
      "entropy": 5.642549324035644,
      "epoch": 0.31460134486071084,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499459959535695,
      "loss": 5.5132,
      "mean_token_accuracy": 0.18159783780574798,
      "num_tokens": 7499046.0,
      "step": 3275
    },
    {
      "entropy": 5.616050767898559,
      "epoch": 0.3150816522574448,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994575830323962,
      "loss": 5.4514,
      "mean_token_accuracy": 0.17918917685747146,
      "num_tokens": 7509853.0,
      "step": 3280
    },
    {
      "entropy": 5.551430940628052,
      "epoch": 0.31556195965417866,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994552013178586,
      "loss": 5.3068,
      "mean_token_accuracy": 0.18969039022922515,
      "num_tokens": 7521091.0,
      "step": 3285
    },
    {
      "entropy": 5.587959337234497,
      "epoch": 0.3160422670509126,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499452814392138,
      "loss": 5.4236,
      "mean_token_accuracy": 0.1940646082162857,
      "num_tokens": 7531317.0,
      "step": 3290
    },
    {
      "entropy": 5.627415990829467,
      "epoch": 0.3165225744476465,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994504222552894,
      "loss": 5.5672,
      "mean_token_accuracy": 0.1813266508281231,
      "num_tokens": 7542822.0,
      "step": 3295
    },
    {
      "entropy": 5.725085687637329,
      "epoch": 0.3170028818443804,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994480249073684,
      "loss": 5.5085,
      "mean_token_accuracy": 0.17990946918725967,
      "num_tokens": 7552434.0,
      "step": 3300
    },
    {
      "entropy": 5.583597040176391,
      "epoch": 0.3174831892411143,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994456223484308,
      "loss": 5.3771,
      "mean_token_accuracy": 0.19330264627933502,
      "num_tokens": 7563895.0,
      "step": 3305
    },
    {
      "entropy": 5.4530237197875975,
      "epoch": 0.31796349663784823,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994432145785323,
      "loss": 5.3979,
      "mean_token_accuracy": 0.18980606347322465,
      "num_tokens": 7575391.0,
      "step": 3310
    },
    {
      "entropy": 5.681956958770752,
      "epoch": 0.3184438040345821,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994408015977288,
      "loss": 5.5493,
      "mean_token_accuracy": 0.19029652178287507,
      "num_tokens": 7587119.0,
      "step": 3315
    },
    {
      "entropy": 5.655573844909668,
      "epoch": 0.31892411143131605,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994383834060764,
      "loss": 5.5209,
      "mean_token_accuracy": 0.17941325157880783,
      "num_tokens": 7598615.0,
      "step": 3320
    },
    {
      "entropy": 5.637982654571533,
      "epoch": 0.31940441882804993,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994359600036311,
      "loss": 5.4826,
      "mean_token_accuracy": 0.19178740084171295,
      "num_tokens": 7610159.0,
      "step": 3325
    },
    {
      "entropy": 5.707050704956055,
      "epoch": 0.31988472622478387,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994335313904493,
      "loss": 5.4699,
      "mean_token_accuracy": 0.18814541697502135,
      "num_tokens": 7620922.0,
      "step": 3330
    },
    {
      "entropy": 5.567379283905029,
      "epoch": 0.32036503362151775,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004994310975665873,
      "loss": 5.386,
      "mean_token_accuracy": 0.19375266283750534,
      "num_tokens": 7632343.0,
      "step": 3335
    },
    {
      "entropy": 5.705099582672119,
      "epoch": 0.3208453410182517,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994286585321017,
      "loss": 5.5929,
      "mean_token_accuracy": 0.1719297081232071,
      "num_tokens": 7644748.0,
      "step": 3340
    },
    {
      "entropy": 5.70050573348999,
      "epoch": 0.32132564841498557,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499426214287049,
      "loss": 5.5325,
      "mean_token_accuracy": 0.19119921773672105,
      "num_tokens": 7655449.0,
      "step": 3345
    },
    {
      "entropy": 5.643219518661499,
      "epoch": 0.3218059558117195,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994237648314862,
      "loss": 5.506,
      "mean_token_accuracy": 0.1850568026304245,
      "num_tokens": 7665623.0,
      "step": 3350
    },
    {
      "entropy": 5.587236118316651,
      "epoch": 0.3222862632084534,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994213101654697,
      "loss": 5.3992,
      "mean_token_accuracy": 0.18939037919044494,
      "num_tokens": 7676860.0,
      "step": 3355
    },
    {
      "entropy": 5.712094449996949,
      "epoch": 0.3227665706051873,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499418850289057,
      "loss": 5.5746,
      "mean_token_accuracy": 0.18135148137807847,
      "num_tokens": 7687778.0,
      "step": 3360
    },
    {
      "entropy": 5.674486255645752,
      "epoch": 0.32324687800192126,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994163852023048,
      "loss": 5.4749,
      "mean_token_accuracy": 0.18185528069734574,
      "num_tokens": 7699154.0,
      "step": 3365
    },
    {
      "entropy": 5.571884107589722,
      "epoch": 0.32372718539865514,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994139149052706,
      "loss": 5.481,
      "mean_token_accuracy": 0.1896562933921814,
      "num_tokens": 7711010.0,
      "step": 3370
    },
    {
      "entropy": 5.700831890106201,
      "epoch": 0.3242074927953891,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994114393980117,
      "loss": 5.5075,
      "mean_token_accuracy": 0.18210598528385163,
      "num_tokens": 7721969.0,
      "step": 3375
    },
    {
      "entropy": 5.675892448425293,
      "epoch": 0.32468780019212296,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994089586805856,
      "loss": 5.4505,
      "mean_token_accuracy": 0.189385986328125,
      "num_tokens": 7733762.0,
      "step": 3380
    },
    {
      "entropy": 5.599011373519898,
      "epoch": 0.3251681075888569,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994064727530496,
      "loss": 5.4692,
      "mean_token_accuracy": 0.18583348244428635,
      "num_tokens": 7744614.0,
      "step": 3385
    },
    {
      "entropy": 5.595316648483276,
      "epoch": 0.3256484149855908,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994039816154618,
      "loss": 5.4031,
      "mean_token_accuracy": 0.19141158014535903,
      "num_tokens": 7755799.0,
      "step": 3390
    },
    {
      "entropy": 5.563256788253784,
      "epoch": 0.3261287223823247,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00049940148526788,
      "loss": 5.4626,
      "mean_token_accuracy": 0.1855825215578079,
      "num_tokens": 7768140.0,
      "step": 3395
    },
    {
      "entropy": 5.60364294052124,
      "epoch": 0.3266090297790586,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993989837103618,
      "loss": 5.4585,
      "mean_token_accuracy": 0.19093261510133744,
      "num_tokens": 7778494.0,
      "step": 3400
    },
    {
      "entropy": 5.65981650352478,
      "epoch": 0.3270893371757925,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993964769429657,
      "loss": 5.5256,
      "mean_token_accuracy": 0.18769457191228867,
      "num_tokens": 7789234.0,
      "step": 3405
    },
    {
      "entropy": 5.611981296539307,
      "epoch": 0.3275696445725264,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993939649657498,
      "loss": 5.5214,
      "mean_token_accuracy": 0.18589239120483397,
      "num_tokens": 7800602.0,
      "step": 3410
    },
    {
      "entropy": 5.560681581497192,
      "epoch": 0.32804995196926034,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004993914477787721,
      "loss": 5.3312,
      "mean_token_accuracy": 0.1929910495877266,
      "num_tokens": 7812803.0,
      "step": 3415
    },
    {
      "entropy": 5.772423648834229,
      "epoch": 0.3285302593659942,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993889253820915,
      "loss": 5.6264,
      "mean_token_accuracy": 0.17467853426933289,
      "num_tokens": 7825432.0,
      "step": 3420
    },
    {
      "entropy": 5.597874021530151,
      "epoch": 0.32901056676272816,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993863977757663,
      "loss": 5.4575,
      "mean_token_accuracy": 0.18343233615159987,
      "num_tokens": 7837258.0,
      "step": 3425
    },
    {
      "entropy": 5.565313911437988,
      "epoch": 0.32949087415946204,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993838649598552,
      "loss": 5.3389,
      "mean_token_accuracy": 0.19879245907068252,
      "num_tokens": 7847573.0,
      "step": 3430
    },
    {
      "entropy": 5.640517377853394,
      "epoch": 0.329971181556196,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993813269344171,
      "loss": 5.4725,
      "mean_token_accuracy": 0.1822481170296669,
      "num_tokens": 7857957.0,
      "step": 3435
    },
    {
      "entropy": 5.587132692337036,
      "epoch": 0.33045148895292986,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993787836995108,
      "loss": 5.3933,
      "mean_token_accuracy": 0.19951517283916473,
      "num_tokens": 7867996.0,
      "step": 3440
    },
    {
      "entropy": 5.546762084960937,
      "epoch": 0.3309317963496638,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993762352551954,
      "loss": 5.4531,
      "mean_token_accuracy": 0.18115595430135728,
      "num_tokens": 7879245.0,
      "step": 3445
    },
    {
      "entropy": 5.667378091812134,
      "epoch": 0.3314121037463977,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993736816015301,
      "loss": 5.531,
      "mean_token_accuracy": 0.17801956832408905,
      "num_tokens": 7891186.0,
      "step": 3450
    },
    {
      "entropy": 5.718769073486328,
      "epoch": 0.3318924111431316,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993711227385742,
      "loss": 5.5526,
      "mean_token_accuracy": 0.18821173906326294,
      "num_tokens": 7902231.0,
      "step": 3455
    },
    {
      "entropy": 5.631023216247558,
      "epoch": 0.3323727185398655,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993685586663871,
      "loss": 5.5036,
      "mean_token_accuracy": 0.187532676756382,
      "num_tokens": 7913364.0,
      "step": 3460
    },
    {
      "entropy": 5.843285274505615,
      "epoch": 0.33285302593659943,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993659893850281,
      "loss": 5.6988,
      "mean_token_accuracy": 0.1718384563922882,
      "num_tokens": 7925217.0,
      "step": 3465
    },
    {
      "entropy": 5.5377867221832275,
      "epoch": 0.3333333333333333,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993634148945573,
      "loss": 5.4338,
      "mean_token_accuracy": 0.18306865096092223,
      "num_tokens": 7937636.0,
      "step": 3470
    },
    {
      "entropy": 5.622658777236938,
      "epoch": 0.33381364073006725,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993608351950341,
      "loss": 5.4672,
      "mean_token_accuracy": 0.17787594795227052,
      "num_tokens": 7948958.0,
      "step": 3475
    },
    {
      "entropy": 5.690193700790405,
      "epoch": 0.33429394812680113,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993582502865185,
      "loss": 5.4938,
      "mean_token_accuracy": 0.18434032946825027,
      "num_tokens": 7960013.0,
      "step": 3480
    },
    {
      "entropy": 5.594874286651612,
      "epoch": 0.33477425552353507,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993556601690706,
      "loss": 5.5073,
      "mean_token_accuracy": 0.18652689158916474,
      "num_tokens": 7971041.0,
      "step": 3485
    },
    {
      "entropy": 5.7146190166473385,
      "epoch": 0.33525456292026895,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993530648427505,
      "loss": 5.5267,
      "mean_token_accuracy": 0.1811326265335083,
      "num_tokens": 7982752.0,
      "step": 3490
    },
    {
      "entropy": 5.6629688262939455,
      "epoch": 0.3357348703170029,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993504643076184,
      "loss": 5.4039,
      "mean_token_accuracy": 0.18894438147544862,
      "num_tokens": 7993681.0,
      "step": 3495
    },
    {
      "entropy": 5.592116165161133,
      "epoch": 0.33621517771373677,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993478585637347,
      "loss": 5.4337,
      "mean_token_accuracy": 0.1875965714454651,
      "num_tokens": 8004727.0,
      "step": 3500
    },
    {
      "entropy": 5.604901647567749,
      "epoch": 0.3366954851104707,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993452476111599,
      "loss": 5.4503,
      "mean_token_accuracy": 0.19188206791877746,
      "num_tokens": 8015423.0,
      "step": 3505
    },
    {
      "entropy": 5.629399013519287,
      "epoch": 0.3371757925072046,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004993426314499546,
      "loss": 5.4163,
      "mean_token_accuracy": 0.19524227529764177,
      "num_tokens": 8027911.0,
      "step": 3510
    },
    {
      "entropy": 5.630470561981201,
      "epoch": 0.3376560999039385,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993400100801796,
      "loss": 5.4435,
      "mean_token_accuracy": 0.19002441614866256,
      "num_tokens": 8038831.0,
      "step": 3515
    },
    {
      "entropy": 5.485448122024536,
      "epoch": 0.3381364073006724,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993373835018956,
      "loss": 5.333,
      "mean_token_accuracy": 0.19800836890935897,
      "num_tokens": 8049906.0,
      "step": 3520
    },
    {
      "entropy": 5.534181308746338,
      "epoch": 0.33861671469740634,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993347517151638,
      "loss": 5.4262,
      "mean_token_accuracy": 0.18656017929315566,
      "num_tokens": 8061158.0,
      "step": 3525
    },
    {
      "entropy": 5.679272985458374,
      "epoch": 0.3390970220941403,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993321147200452,
      "loss": 5.4049,
      "mean_token_accuracy": 0.18500697761774063,
      "num_tokens": 8071958.0,
      "step": 3530
    },
    {
      "entropy": 5.590602445602417,
      "epoch": 0.33957732949087416,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499329472516601,
      "loss": 5.4023,
      "mean_token_accuracy": 0.1830559030175209,
      "num_tokens": 8084068.0,
      "step": 3535
    },
    {
      "entropy": 5.501382446289062,
      "epoch": 0.3400576368876081,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993268251048925,
      "loss": 5.3107,
      "mean_token_accuracy": 0.2009879857301712,
      "num_tokens": 8096132.0,
      "step": 3540
    },
    {
      "entropy": 5.591581535339356,
      "epoch": 0.340537944284342,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993241724849814,
      "loss": 5.4852,
      "mean_token_accuracy": 0.18972329199314117,
      "num_tokens": 8107327.0,
      "step": 3545
    },
    {
      "entropy": 5.586108875274658,
      "epoch": 0.3410182516810759,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499321514656929,
      "loss": 5.451,
      "mean_token_accuracy": 0.18751484006643296,
      "num_tokens": 8118584.0,
      "step": 3550
    },
    {
      "entropy": 5.632966232299805,
      "epoch": 0.3414985590778098,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993188516207972,
      "loss": 5.4978,
      "mean_token_accuracy": 0.18059134632349014,
      "num_tokens": 8130081.0,
      "step": 3555
    },
    {
      "entropy": 5.603939819335937,
      "epoch": 0.34197886647454373,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993161833766478,
      "loss": 5.4545,
      "mean_token_accuracy": 0.18668154925107955,
      "num_tokens": 8141463.0,
      "step": 3560
    },
    {
      "entropy": 5.662025403976441,
      "epoch": 0.3424591738712776,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993135099245426,
      "loss": 5.4714,
      "mean_token_accuracy": 0.18695254474878312,
      "num_tokens": 8153863.0,
      "step": 3565
    },
    {
      "entropy": 5.524808263778686,
      "epoch": 0.34293948126801155,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993108312645438,
      "loss": 5.423,
      "mean_token_accuracy": 0.18549051284790039,
      "num_tokens": 8165695.0,
      "step": 3570
    },
    {
      "entropy": 5.607334613800049,
      "epoch": 0.34341978866474543,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993081473967135,
      "loss": 5.4845,
      "mean_token_accuracy": 0.1852959141135216,
      "num_tokens": 8176456.0,
      "step": 3575
    },
    {
      "entropy": 5.65879111289978,
      "epoch": 0.34390009606147937,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993054583211143,
      "loss": 5.4549,
      "mean_token_accuracy": 0.19000289142131804,
      "num_tokens": 8189050.0,
      "step": 3580
    },
    {
      "entropy": 5.562312746047974,
      "epoch": 0.34438040345821325,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004993027640378081,
      "loss": 5.3878,
      "mean_token_accuracy": 0.18902230262756348,
      "num_tokens": 8200011.0,
      "step": 3585
    },
    {
      "entropy": 5.573532772064209,
      "epoch": 0.3448607108549472,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499300064546858,
      "loss": 5.3829,
      "mean_token_accuracy": 0.19188244193792342,
      "num_tokens": 8211770.0,
      "step": 3590
    },
    {
      "entropy": 5.589154958724976,
      "epoch": 0.34534101825168106,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992973598483264,
      "loss": 5.4252,
      "mean_token_accuracy": 0.19337030947208406,
      "num_tokens": 8223582.0,
      "step": 3595
    },
    {
      "entropy": 5.698397827148438,
      "epoch": 0.345821325648415,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499294649942276,
      "loss": 5.553,
      "mean_token_accuracy": 0.18515172749757766,
      "num_tokens": 8234336.0,
      "step": 3600
    },
    {
      "entropy": 5.645198345184326,
      "epoch": 0.3463016330451489,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992919348287699,
      "loss": 5.4546,
      "mean_token_accuracy": 0.187334606051445,
      "num_tokens": 8244605.0,
      "step": 3605
    },
    {
      "entropy": 5.615684843063354,
      "epoch": 0.3467819404418828,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004992892145078711,
      "loss": 5.5042,
      "mean_token_accuracy": 0.18244509696960448,
      "num_tokens": 8255876.0,
      "step": 3610
    },
    {
      "entropy": 5.569511413574219,
      "epoch": 0.3472622478386167,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992864889796427,
      "loss": 5.3968,
      "mean_token_accuracy": 0.19007992446422578,
      "num_tokens": 8266602.0,
      "step": 3615
    },
    {
      "entropy": 5.599522733688355,
      "epoch": 0.34774255523535064,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992837582441481,
      "loss": 5.38,
      "mean_token_accuracy": 0.1920741856098175,
      "num_tokens": 8279804.0,
      "step": 3620
    },
    {
      "entropy": 5.669801664352417,
      "epoch": 0.3482228626320845,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992810223014506,
      "loss": 5.4916,
      "mean_token_accuracy": 0.18763836473226547,
      "num_tokens": 8291020.0,
      "step": 3625
    },
    {
      "entropy": 5.6229675769805905,
      "epoch": 0.34870317002881845,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992782811516137,
      "loss": 5.4266,
      "mean_token_accuracy": 0.19148036688566208,
      "num_tokens": 8302192.0,
      "step": 3630
    },
    {
      "entropy": 5.604986190795898,
      "epoch": 0.34918347742555234,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004992755347947011,
      "loss": 5.4054,
      "mean_token_accuracy": 0.1876965194940567,
      "num_tokens": 8313649.0,
      "step": 3635
    },
    {
      "entropy": 5.560658979415893,
      "epoch": 0.34966378482228627,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992727832307766,
      "loss": 5.4003,
      "mean_token_accuracy": 0.19191381931304932,
      "num_tokens": 8324694.0,
      "step": 3640
    },
    {
      "entropy": 5.679537630081176,
      "epoch": 0.35014409221902015,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992700264599039,
      "loss": 5.5466,
      "mean_token_accuracy": 0.17840450406074523,
      "num_tokens": 8336517.0,
      "step": 3645
    },
    {
      "entropy": 5.643055534362793,
      "epoch": 0.3506243996157541,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992672644821473,
      "loss": 5.4974,
      "mean_token_accuracy": 0.18391885459423066,
      "num_tokens": 8349001.0,
      "step": 3650
    },
    {
      "entropy": 5.640076971054077,
      "epoch": 0.35110470701248797,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992644972975707,
      "loss": 5.4248,
      "mean_token_accuracy": 0.18910549581050873,
      "num_tokens": 8361230.0,
      "step": 3655
    },
    {
      "entropy": 5.5414430618286135,
      "epoch": 0.3515850144092219,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992617249062383,
      "loss": 5.3548,
      "mean_token_accuracy": 0.19453334510326387,
      "num_tokens": 8372159.0,
      "step": 3660
    },
    {
      "entropy": 5.589181232452392,
      "epoch": 0.3520653218059558,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992589473082147,
      "loss": 5.4684,
      "mean_token_accuracy": 0.18958093523979186,
      "num_tokens": 8383228.0,
      "step": 3665
    },
    {
      "entropy": 5.648845243453979,
      "epoch": 0.3525456292026897,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992561645035641,
      "loss": 5.4072,
      "mean_token_accuracy": 0.18761504739522933,
      "num_tokens": 8394582.0,
      "step": 3670
    },
    {
      "entropy": 5.605154609680175,
      "epoch": 0.3530259365994236,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992533764923515,
      "loss": 5.4197,
      "mean_token_accuracy": 0.18835971355438233,
      "num_tokens": 8406784.0,
      "step": 3675
    },
    {
      "entropy": 5.585361003875732,
      "epoch": 0.35350624399615754,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992505832746412,
      "loss": 5.4038,
      "mean_token_accuracy": 0.19672708660364152,
      "num_tokens": 8418405.0,
      "step": 3680
    },
    {
      "entropy": 5.5960245609283445,
      "epoch": 0.3539865513928914,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992477848504983,
      "loss": 5.3541,
      "mean_token_accuracy": 0.18893332481384278,
      "num_tokens": 8430432.0,
      "step": 3685
    },
    {
      "entropy": 5.607988166809082,
      "epoch": 0.35446685878962536,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004992449812199877,
      "loss": 5.5037,
      "mean_token_accuracy": 0.18349802643060684,
      "num_tokens": 8442423.0,
      "step": 3690
    },
    {
      "entropy": 5.622403383255005,
      "epoch": 0.3549471661863593,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992421723831745,
      "loss": 5.5182,
      "mean_token_accuracy": 0.18986132144927978,
      "num_tokens": 8454951.0,
      "step": 3695
    },
    {
      "entropy": 5.564925622940064,
      "epoch": 0.3554274735830932,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004992393583401239,
      "loss": 5.3553,
      "mean_token_accuracy": 0.19411883056163787,
      "num_tokens": 8467758.0,
      "step": 3700
    },
    {
      "entropy": 5.60244460105896,
      "epoch": 0.3559077809798271,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992365390909014,
      "loss": 5.4398,
      "mean_token_accuracy": 0.18677409291267394,
      "num_tokens": 8479728.0,
      "step": 3705
    },
    {
      "entropy": 5.604581356048584,
      "epoch": 0.356388088376561,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992337146355721,
      "loss": 5.5095,
      "mean_token_accuracy": 0.18320033997297286,
      "num_tokens": 8492099.0,
      "step": 3710
    },
    {
      "entropy": 5.657941150665283,
      "epoch": 0.35686839577329493,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992308849742019,
      "loss": 5.4536,
      "mean_token_accuracy": 0.17795021831989288,
      "num_tokens": 8504657.0,
      "step": 3715
    },
    {
      "entropy": 5.625857305526734,
      "epoch": 0.3573487031700288,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992280501068563,
      "loss": 5.4227,
      "mean_token_accuracy": 0.18991409093141556,
      "num_tokens": 8514728.0,
      "step": 3720
    },
    {
      "entropy": 5.62817611694336,
      "epoch": 0.35782901056676275,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992252100336012,
      "loss": 5.5363,
      "mean_token_accuracy": 0.18928872793912888,
      "num_tokens": 8525588.0,
      "step": 3725
    },
    {
      "entropy": 5.618461799621582,
      "epoch": 0.35830931796349663,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992223647545027,
      "loss": 5.4997,
      "mean_token_accuracy": 0.18794921934604644,
      "num_tokens": 8537468.0,
      "step": 3730
    },
    {
      "entropy": 5.642933511734009,
      "epoch": 0.35878962536023057,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992195142696266,
      "loss": 5.4099,
      "mean_token_accuracy": 0.1922216534614563,
      "num_tokens": 8548598.0,
      "step": 3735
    },
    {
      "entropy": 5.412218189239502,
      "epoch": 0.35926993275696445,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992166585790391,
      "loss": 5.2967,
      "mean_token_accuracy": 0.20350995361804963,
      "num_tokens": 8560301.0,
      "step": 3740
    },
    {
      "entropy": 5.571508169174194,
      "epoch": 0.3597502401536984,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992137976828067,
      "loss": 5.3904,
      "mean_token_accuracy": 0.19346483945846557,
      "num_tokens": 8571186.0,
      "step": 3745
    },
    {
      "entropy": 5.607302665710449,
      "epoch": 0.36023054755043227,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992109315809955,
      "loss": 5.3967,
      "mean_token_accuracy": 0.19522017240524292,
      "num_tokens": 8580725.0,
      "step": 3750
    },
    {
      "entropy": 5.5793187618255615,
      "epoch": 0.3607108549471662,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004992080602736725,
      "loss": 5.5152,
      "mean_token_accuracy": 0.18237872421741486,
      "num_tokens": 8594598.0,
      "step": 3755
    },
    {
      "entropy": 5.718835163116455,
      "epoch": 0.3611911623439001,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004992051837609039,
      "loss": 5.5138,
      "mean_token_accuracy": 0.18184855878353118,
      "num_tokens": 8606733.0,
      "step": 3760
    },
    {
      "entropy": 5.649420595169067,
      "epoch": 0.361671469740634,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992023020427568,
      "loss": 5.4442,
      "mean_token_accuracy": 0.1857554316520691,
      "num_tokens": 8618863.0,
      "step": 3765
    },
    {
      "entropy": 5.453961610794067,
      "epoch": 0.3621517771373679,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991994151192979,
      "loss": 5.2943,
      "mean_token_accuracy": 0.1941133290529251,
      "num_tokens": 8629270.0,
      "step": 3770
    },
    {
      "entropy": 5.610463285446167,
      "epoch": 0.36263208453410184,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991965229905943,
      "loss": 5.4909,
      "mean_token_accuracy": 0.19132499098777772,
      "num_tokens": 8641363.0,
      "step": 3775
    },
    {
      "entropy": 5.66168212890625,
      "epoch": 0.3631123919308357,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991936256567133,
      "loss": 5.4499,
      "mean_token_accuracy": 0.1907256633043289,
      "num_tokens": 8653233.0,
      "step": 3780
    },
    {
      "entropy": 5.571015691757202,
      "epoch": 0.36359269932756966,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000499190723117722,
      "loss": 5.4484,
      "mean_token_accuracy": 0.18239522725343704,
      "num_tokens": 8665192.0,
      "step": 3785
    },
    {
      "entropy": 5.669944858551025,
      "epoch": 0.36407300672430354,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991878153736877,
      "loss": 5.527,
      "mean_token_accuracy": 0.181233774125576,
      "num_tokens": 8677669.0,
      "step": 3790
    },
    {
      "entropy": 5.529652452468872,
      "epoch": 0.3645533141210375,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991849024246781,
      "loss": 5.3266,
      "mean_token_accuracy": 0.19549092054367065,
      "num_tokens": 8688002.0,
      "step": 3795
    },
    {
      "entropy": 5.552938508987427,
      "epoch": 0.36503362151777136,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991819842707608,
      "loss": 5.3809,
      "mean_token_accuracy": 0.19497113823890685,
      "num_tokens": 8698396.0,
      "step": 3800
    },
    {
      "entropy": 5.656788778305054,
      "epoch": 0.3655139289145053,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991790609120035,
      "loss": 5.3855,
      "mean_token_accuracy": 0.19022500067949294,
      "num_tokens": 8711135.0,
      "step": 3805
    },
    {
      "entropy": 5.545717477798462,
      "epoch": 0.3659942363112392,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499176132348474,
      "loss": 5.4329,
      "mean_token_accuracy": 0.19408364295959474,
      "num_tokens": 8723707.0,
      "step": 3810
    },
    {
      "entropy": 5.6224829196929935,
      "epoch": 0.3664745437079731,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991731985802405,
      "loss": 5.3788,
      "mean_token_accuracy": 0.19512805342674255,
      "num_tokens": 8734193.0,
      "step": 3815
    },
    {
      "entropy": 5.575182294845581,
      "epoch": 0.366954851104707,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991702596073708,
      "loss": 5.4498,
      "mean_token_accuracy": 0.18577555567026138,
      "num_tokens": 8745619.0,
      "step": 3820
    },
    {
      "entropy": 5.521268081665039,
      "epoch": 0.36743515850144093,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991673154299335,
      "loss": 5.3718,
      "mean_token_accuracy": 0.18944036811590195,
      "num_tokens": 8757331.0,
      "step": 3825
    },
    {
      "entropy": 5.584135866165161,
      "epoch": 0.3679154658981748,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991643660479967,
      "loss": 5.3848,
      "mean_token_accuracy": 0.19181567877531053,
      "num_tokens": 8768840.0,
      "step": 3830
    },
    {
      "entropy": 5.572615623474121,
      "epoch": 0.36839577329490875,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991614114616289,
      "loss": 5.3373,
      "mean_token_accuracy": 0.19323140680789946,
      "num_tokens": 8781214.0,
      "step": 3835
    },
    {
      "entropy": 5.598087167739868,
      "epoch": 0.3688760806916426,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004991584516708988,
      "loss": 5.407,
      "mean_token_accuracy": 0.18879638165235518,
      "num_tokens": 8791645.0,
      "step": 3840
    },
    {
      "entropy": 5.666203022003174,
      "epoch": 0.36935638808837656,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991554866758751,
      "loss": 5.58,
      "mean_token_accuracy": 0.17914286777377128,
      "num_tokens": 8803286.0,
      "step": 3845
    },
    {
      "entropy": 5.6028059959411625,
      "epoch": 0.36983669548511044,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991525164766265,
      "loss": 5.3582,
      "mean_token_accuracy": 0.19065531343221664,
      "num_tokens": 8814207.0,
      "step": 3850
    },
    {
      "entropy": 5.574209070205688,
      "epoch": 0.3703170028818444,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004991495410732222,
      "loss": 5.4283,
      "mean_token_accuracy": 0.18504399955272674,
      "num_tokens": 8825540.0,
      "step": 3855
    },
    {
      "entropy": 5.650593423843384,
      "epoch": 0.37079731027857826,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991465604657311,
      "loss": 5.5504,
      "mean_token_accuracy": 0.17896147668361664,
      "num_tokens": 8838182.0,
      "step": 3860
    },
    {
      "entropy": 5.579078006744385,
      "epoch": 0.3712776176753122,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991435746542224,
      "loss": 5.4175,
      "mean_token_accuracy": 0.1951797842979431,
      "num_tokens": 8850211.0,
      "step": 3865
    },
    {
      "entropy": 5.548032665252686,
      "epoch": 0.37175792507204614,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991405836387655,
      "loss": 5.4724,
      "mean_token_accuracy": 0.18146964982151986,
      "num_tokens": 8862804.0,
      "step": 3870
    },
    {
      "entropy": 5.588915157318115,
      "epoch": 0.37223823246878,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991375874194298,
      "loss": 5.4164,
      "mean_token_accuracy": 0.1890851303935051,
      "num_tokens": 8874112.0,
      "step": 3875
    },
    {
      "entropy": 5.566713428497314,
      "epoch": 0.37271853986551395,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499134585996285,
      "loss": 5.4356,
      "mean_token_accuracy": 0.18930822014808654,
      "num_tokens": 8885114.0,
      "step": 3880
    },
    {
      "entropy": 5.597771644592285,
      "epoch": 0.37319884726224783,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991315793694004,
      "loss": 5.329,
      "mean_token_accuracy": 0.1946409046649933,
      "num_tokens": 8895555.0,
      "step": 3885
    },
    {
      "entropy": 5.520039701461792,
      "epoch": 0.37367915465898177,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991285675388463,
      "loss": 5.3281,
      "mean_token_accuracy": 0.20043965727090834,
      "num_tokens": 8906073.0,
      "step": 3890
    },
    {
      "entropy": 5.59633960723877,
      "epoch": 0.37415946205571565,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991255505046922,
      "loss": 5.4848,
      "mean_token_accuracy": 0.1856019526720047,
      "num_tokens": 8916587.0,
      "step": 3895
    },
    {
      "entropy": 5.615598583221436,
      "epoch": 0.3746397694524496,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991225282670083,
      "loss": 5.3741,
      "mean_token_accuracy": 0.1854495272040367,
      "num_tokens": 8927923.0,
      "step": 3900
    },
    {
      "entropy": 5.522138500213623,
      "epoch": 0.37512007684918347,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499119500825865,
      "loss": 5.4171,
      "mean_token_accuracy": 0.19048261940479277,
      "num_tokens": 8939939.0,
      "step": 3905
    },
    {
      "entropy": 5.522910451889038,
      "epoch": 0.3756003842459174,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991164681813323,
      "loss": 5.4,
      "mean_token_accuracy": 0.19586212038993836,
      "num_tokens": 8951748.0,
      "step": 3910
    },
    {
      "entropy": 5.554461240768433,
      "epoch": 0.3760806916426513,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991134303334807,
      "loss": 5.32,
      "mean_token_accuracy": 0.19885835498571397,
      "num_tokens": 8962922.0,
      "step": 3915
    },
    {
      "entropy": 5.472480392456054,
      "epoch": 0.3765609990393852,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991103872823807,
      "loss": 5.2987,
      "mean_token_accuracy": 0.2007770374417305,
      "num_tokens": 8974013.0,
      "step": 3920
    },
    {
      "entropy": 5.536104154586792,
      "epoch": 0.3770413064361191,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499107339028103,
      "loss": 5.377,
      "mean_token_accuracy": 0.18797796219587326,
      "num_tokens": 8986032.0,
      "step": 3925
    },
    {
      "entropy": 5.624913358688355,
      "epoch": 0.37752161383285304,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991042855707184,
      "loss": 5.3747,
      "mean_token_accuracy": 0.18829465806484222,
      "num_tokens": 8996889.0,
      "step": 3930
    },
    {
      "entropy": 5.5282268047332765,
      "epoch": 0.3780019212295869,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991012269102977,
      "loss": 5.3596,
      "mean_token_accuracy": 0.19118607342243193,
      "num_tokens": 9007594.0,
      "step": 3935
    },
    {
      "entropy": 5.473355913162232,
      "epoch": 0.37848222862632086,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990981630469119,
      "loss": 5.3455,
      "mean_token_accuracy": 0.18814336955547334,
      "num_tokens": 9018097.0,
      "step": 3940
    },
    {
      "entropy": 5.59659333229065,
      "epoch": 0.37896253602305474,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990950939806323,
      "loss": 5.4724,
      "mean_token_accuracy": 0.18152033388614655,
      "num_tokens": 9029554.0,
      "step": 3945
    },
    {
      "entropy": 5.562332487106323,
      "epoch": 0.3794428434197887,
      "grad_norm": 0.859375,
      "learning_rate": 0.00049909201971153,
      "loss": 5.3375,
      "mean_token_accuracy": 0.19086694419384004,
      "num_tokens": 9042518.0,
      "step": 3950
    },
    {
      "entropy": 5.520054674148559,
      "epoch": 0.37992315081652256,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990889402396763,
      "loss": 5.3912,
      "mean_token_accuracy": 0.19118970036506652,
      "num_tokens": 9054524.0,
      "step": 3955
    },
    {
      "entropy": 5.604040002822876,
      "epoch": 0.3804034582132565,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990858555651431,
      "loss": 5.3534,
      "mean_token_accuracy": 0.19073721766471863,
      "num_tokens": 9065375.0,
      "step": 3960
    },
    {
      "entropy": 5.519604921340942,
      "epoch": 0.3808837656099904,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990827656880015,
      "loss": 5.3067,
      "mean_token_accuracy": 0.19381832629442214,
      "num_tokens": 9076338.0,
      "step": 3965
    },
    {
      "entropy": 5.574223232269287,
      "epoch": 0.3813640730067243,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990796706083235,
      "loss": 5.3524,
      "mean_token_accuracy": 0.19108670055866242,
      "num_tokens": 9088407.0,
      "step": 3970
    },
    {
      "entropy": 5.517913103103638,
      "epoch": 0.3818443804034582,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990765703261809,
      "loss": 5.3182,
      "mean_token_accuracy": 0.19173867255449295,
      "num_tokens": 9099833.0,
      "step": 3975
    },
    {
      "entropy": 5.397404193878174,
      "epoch": 0.38232468780019213,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004990734648416458,
      "loss": 5.2948,
      "mean_token_accuracy": 0.19632792323827744,
      "num_tokens": 9111126.0,
      "step": 3980
    },
    {
      "entropy": 5.605161476135254,
      "epoch": 0.382804995196926,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990703541547901,
      "loss": 5.3929,
      "mean_token_accuracy": 0.19436297714710235,
      "num_tokens": 9121979.0,
      "step": 3985
    },
    {
      "entropy": 5.5903404712677,
      "epoch": 0.38328530259365995,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990672382656863,
      "loss": 5.4058,
      "mean_token_accuracy": 0.19064545035362243,
      "num_tokens": 9132929.0,
      "step": 3990
    },
    {
      "entropy": 5.596808624267578,
      "epoch": 0.38376560999039383,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990641171744064,
      "loss": 5.3732,
      "mean_token_accuracy": 0.1942216232419014,
      "num_tokens": 9143903.0,
      "step": 3995
    },
    {
      "entropy": 5.588313817977905,
      "epoch": 0.38424591738712777,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990609908810231,
      "loss": 5.4502,
      "mean_token_accuracy": 0.18742062598466874,
      "num_tokens": 9154416.0,
      "step": 4000
    },
    {
      "entropy": 5.582321405410767,
      "epoch": 0.38472622478386165,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990578593856089,
      "loss": 5.4266,
      "mean_token_accuracy": 0.18962429016828536,
      "num_tokens": 9165613.0,
      "step": 4005
    },
    {
      "entropy": 5.491159725189209,
      "epoch": 0.3852065321805956,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990547226882366,
      "loss": 5.3813,
      "mean_token_accuracy": 0.19434286206960677,
      "num_tokens": 9177884.0,
      "step": 4010
    },
    {
      "entropy": 5.613833093643189,
      "epoch": 0.38568683957732947,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004990515807889788,
      "loss": 5.5091,
      "mean_token_accuracy": 0.18056989163160325,
      "num_tokens": 9190041.0,
      "step": 4015
    },
    {
      "entropy": 5.628996181488037,
      "epoch": 0.3861671469740634,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990484336879087,
      "loss": 5.3944,
      "mean_token_accuracy": 0.19238094687461854,
      "num_tokens": 9202390.0,
      "step": 4020
    },
    {
      "entropy": 5.512400102615357,
      "epoch": 0.3866474543707973,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990452813850992,
      "loss": 5.3972,
      "mean_token_accuracy": 0.1906796544790268,
      "num_tokens": 9213437.0,
      "step": 4025
    },
    {
      "entropy": 5.599765205383301,
      "epoch": 0.3871277617675312,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990421238806236,
      "loss": 5.4618,
      "mean_token_accuracy": 0.18272736817598342,
      "num_tokens": 9226310.0,
      "step": 4030
    },
    {
      "entropy": 5.603873920440674,
      "epoch": 0.38760806916426516,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990389611745551,
      "loss": 5.4077,
      "mean_token_accuracy": 0.19082859754562378,
      "num_tokens": 9236271.0,
      "step": 4035
    },
    {
      "entropy": 5.613268089294434,
      "epoch": 0.38808837656099904,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990357932669672,
      "loss": 5.4624,
      "mean_token_accuracy": 0.19046298116445542,
      "num_tokens": 9247755.0,
      "step": 4040
    },
    {
      "entropy": 5.542822551727295,
      "epoch": 0.388568683957733,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990326201579335,
      "loss": 5.3169,
      "mean_token_accuracy": 0.20392679125070573,
      "num_tokens": 9259821.0,
      "step": 4045
    },
    {
      "entropy": 5.564394140243531,
      "epoch": 0.38904899135446686,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990294418475274,
      "loss": 5.4153,
      "mean_token_accuracy": 0.18946493715047835,
      "num_tokens": 9270663.0,
      "step": 4050
    },
    {
      "entropy": 5.552659273147583,
      "epoch": 0.3895292987512008,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004990262583358231,
      "loss": 5.4273,
      "mean_token_accuracy": 0.1895113468170166,
      "num_tokens": 9282588.0,
      "step": 4055
    },
    {
      "entropy": 5.5941455364227295,
      "epoch": 0.3900096061479347,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990230696228943,
      "loss": 5.4071,
      "mean_token_accuracy": 0.17811299264431,
      "num_tokens": 9293368.0,
      "step": 4060
    },
    {
      "entropy": 5.548636865615845,
      "epoch": 0.3904899135446686,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990198757088149,
      "loss": 5.468,
      "mean_token_accuracy": 0.1825200393795967,
      "num_tokens": 9305962.0,
      "step": 4065
    },
    {
      "entropy": 5.558157014846802,
      "epoch": 0.3909702209414025,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990166765936593,
      "loss": 5.3417,
      "mean_token_accuracy": 0.1954346388578415,
      "num_tokens": 9317955.0,
      "step": 4070
    },
    {
      "entropy": 5.549290418624878,
      "epoch": 0.3914505283381364,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990134722775016,
      "loss": 5.3593,
      "mean_token_accuracy": 0.19245800226926804,
      "num_tokens": 9329491.0,
      "step": 4075
    },
    {
      "entropy": 5.498097085952759,
      "epoch": 0.3919308357348703,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990102627604162,
      "loss": 5.4817,
      "mean_token_accuracy": 0.19573085010051727,
      "num_tokens": 9341612.0,
      "step": 4080
    },
    {
      "entropy": 5.5749123096466064,
      "epoch": 0.39241114313160425,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990070480424778,
      "loss": 5.4291,
      "mean_token_accuracy": 0.18815626055002213,
      "num_tokens": 9352302.0,
      "step": 4085
    },
    {
      "entropy": 5.589935445785523,
      "epoch": 0.3928914505283381,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990038281237608,
      "loss": 5.3586,
      "mean_token_accuracy": 0.19239330440759658,
      "num_tokens": 9363303.0,
      "step": 4090
    },
    {
      "entropy": 5.502812719345092,
      "epoch": 0.39337175792507206,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990006030043401,
      "loss": 5.3286,
      "mean_token_accuracy": 0.18687490820884706,
      "num_tokens": 9375878.0,
      "step": 4095
    },
    {
      "entropy": 5.534584331512451,
      "epoch": 0.39385206532180594,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989973726842906,
      "loss": 5.3734,
      "mean_token_accuracy": 0.19118351340293885,
      "num_tokens": 9388342.0,
      "step": 4100
    },
    {
      "entropy": 5.500079584121704,
      "epoch": 0.3943323727185399,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989941371636872,
      "loss": 5.2949,
      "mean_token_accuracy": 0.1912856638431549,
      "num_tokens": 9399047.0,
      "step": 4105
    },
    {
      "entropy": 5.587457847595215,
      "epoch": 0.39481268011527376,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989908964426051,
      "loss": 5.3811,
      "mean_token_accuracy": 0.19598662108182907,
      "num_tokens": 9410172.0,
      "step": 4110
    },
    {
      "entropy": 5.603793287277222,
      "epoch": 0.3952929875120077,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989876505211194,
      "loss": 5.5287,
      "mean_token_accuracy": 0.18236763179302215,
      "num_tokens": 9422287.0,
      "step": 4115
    },
    {
      "entropy": 5.639399194717408,
      "epoch": 0.3957732949087416,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989843993993056,
      "loss": 5.382,
      "mean_token_accuracy": 0.19403408318758011,
      "num_tokens": 9433709.0,
      "step": 4120
    },
    {
      "entropy": 5.453795385360718,
      "epoch": 0.3962536023054755,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989811430772392,
      "loss": 5.2754,
      "mean_token_accuracy": 0.19953108578920364,
      "num_tokens": 9445138.0,
      "step": 4125
    },
    {
      "entropy": 5.477317523956299,
      "epoch": 0.3967339097022094,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989778815549957,
      "loss": 5.4076,
      "mean_token_accuracy": 0.18944294303655623,
      "num_tokens": 9455263.0,
      "step": 4130
    },
    {
      "entropy": 5.614648723602295,
      "epoch": 0.39721421709894333,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989746148326508,
      "loss": 5.3727,
      "mean_token_accuracy": 0.18981866538524628,
      "num_tokens": 9465491.0,
      "step": 4135
    },
    {
      "entropy": 5.482787322998047,
      "epoch": 0.3976945244956772,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004989713429102805,
      "loss": 5.339,
      "mean_token_accuracy": 0.19033195674419404,
      "num_tokens": 9477601.0,
      "step": 4140
    },
    {
      "entropy": 5.545722436904907,
      "epoch": 0.39817483189241115,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004989680657879607,
      "loss": 5.3988,
      "mean_token_accuracy": 0.19300916343927382,
      "num_tokens": 9489385.0,
      "step": 4145
    },
    {
      "entropy": 5.52602744102478,
      "epoch": 0.39865513928914503,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989647834657675,
      "loss": 5.2736,
      "mean_token_accuracy": 0.1982199728488922,
      "num_tokens": 9501131.0,
      "step": 4150
    },
    {
      "entropy": 5.4960432052612305,
      "epoch": 0.39913544668587897,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498961495943777,
      "loss": 5.4284,
      "mean_token_accuracy": 0.19547935873270034,
      "num_tokens": 9513094.0,
      "step": 4155
    },
    {
      "entropy": 5.630834293365479,
      "epoch": 0.39961575408261285,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989582032220656,
      "loss": 5.5256,
      "mean_token_accuracy": 0.1863490089774132,
      "num_tokens": 9524538.0,
      "step": 4160
    },
    {
      "entropy": 5.569158506393433,
      "epoch": 0.4000960614793468,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989549053007096,
      "loss": 5.3489,
      "mean_token_accuracy": 0.20113594233989715,
      "num_tokens": 9535284.0,
      "step": 4165
    },
    {
      "entropy": 5.550960111618042,
      "epoch": 0.40057636887608067,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989516021797858,
      "loss": 5.4361,
      "mean_token_accuracy": 0.1874231815338135,
      "num_tokens": 9546472.0,
      "step": 4170
    },
    {
      "entropy": 5.565934753417968,
      "epoch": 0.4010566762728146,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498948293859371,
      "loss": 5.4279,
      "mean_token_accuracy": 0.18724920451641083,
      "num_tokens": 9558358.0,
      "step": 4175
    },
    {
      "entropy": 5.600048398971557,
      "epoch": 0.4015369836695485,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004989449803395415,
      "loss": 5.4499,
      "mean_token_accuracy": 0.19195451587438583,
      "num_tokens": 9570653.0,
      "step": 4180
    },
    {
      "entropy": 5.564519691467285,
      "epoch": 0.4020172910662824,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989416616203747,
      "loss": 5.3972,
      "mean_token_accuracy": 0.19142527729272843,
      "num_tokens": 9582150.0,
      "step": 4185
    },
    {
      "entropy": 5.590189504623413,
      "epoch": 0.4024975984630163,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989383377019476,
      "loss": 5.3369,
      "mean_token_accuracy": 0.19532157182693483,
      "num_tokens": 9592462.0,
      "step": 4190
    },
    {
      "entropy": 5.4742389678955075,
      "epoch": 0.40297790585975024,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989350085843371,
      "loss": 5.3327,
      "mean_token_accuracy": 0.19161231219768524,
      "num_tokens": 9604027.0,
      "step": 4195
    },
    {
      "entropy": 5.4748578548431395,
      "epoch": 0.4034582132564842,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004989316742676207,
      "loss": 5.3185,
      "mean_token_accuracy": 0.19610172063112258,
      "num_tokens": 9616325.0,
      "step": 4200
    },
    {
      "entropy": 5.458920621871949,
      "epoch": 0.40393852065321806,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989283347518757,
      "loss": 5.2904,
      "mean_token_accuracy": 0.1975201338529587,
      "num_tokens": 9628133.0,
      "step": 4205
    },
    {
      "entropy": 5.664987993240357,
      "epoch": 0.404418828049952,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989249900371797,
      "loss": 5.5003,
      "mean_token_accuracy": 0.18330966383218766,
      "num_tokens": 9639686.0,
      "step": 4210
    },
    {
      "entropy": 5.555954265594482,
      "epoch": 0.4048991354466859,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989216401236103,
      "loss": 5.3865,
      "mean_token_accuracy": 0.1899016246199608,
      "num_tokens": 9650222.0,
      "step": 4215
    },
    {
      "entropy": 5.458903455734253,
      "epoch": 0.4053794428434198,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989182850112455,
      "loss": 5.3048,
      "mean_token_accuracy": 0.20585627257823944,
      "num_tokens": 9661792.0,
      "step": 4220
    },
    {
      "entropy": 5.544176006317139,
      "epoch": 0.4058597502401537,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989149247001629,
      "loss": 5.4053,
      "mean_token_accuracy": 0.18887956887483598,
      "num_tokens": 9673000.0,
      "step": 4225
    },
    {
      "entropy": 5.611917209625244,
      "epoch": 0.40634005763688763,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989115591904407,
      "loss": 5.3492,
      "mean_token_accuracy": 0.19597309231758117,
      "num_tokens": 9685253.0,
      "step": 4230
    },
    {
      "entropy": 5.489803409576416,
      "epoch": 0.4068203650336215,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004989081884821569,
      "loss": 5.3578,
      "mean_token_accuracy": 0.1868518963456154,
      "num_tokens": 9697245.0,
      "step": 4235
    },
    {
      "entropy": 5.496517181396484,
      "epoch": 0.40730067243035545,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989048125753899,
      "loss": 5.3624,
      "mean_token_accuracy": 0.19268081486225128,
      "num_tokens": 9710095.0,
      "step": 4240
    },
    {
      "entropy": 5.51132984161377,
      "epoch": 0.40778097982708933,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498901431470218,
      "loss": 5.2435,
      "mean_token_accuracy": 0.20017803013324736,
      "num_tokens": 9721488.0,
      "step": 4245
    },
    {
      "entropy": 5.369005727767944,
      "epoch": 0.40826128722382327,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004988980451667198,
      "loss": 5.2183,
      "mean_token_accuracy": 0.1994133025407791,
      "num_tokens": 9733280.0,
      "step": 4250
    },
    {
      "entropy": 5.505516052246094,
      "epoch": 0.40874159462055715,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988946536649737,
      "loss": 5.3349,
      "mean_token_accuracy": 0.19074197262525558,
      "num_tokens": 9744514.0,
      "step": 4255
    },
    {
      "entropy": 5.514144372940064,
      "epoch": 0.4092219020172911,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988912569650585,
      "loss": 5.349,
      "mean_token_accuracy": 0.19237666875123977,
      "num_tokens": 9754931.0,
      "step": 4260
    },
    {
      "entropy": 5.476861238479614,
      "epoch": 0.40970220941402496,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988878550670533,
      "loss": 5.3106,
      "mean_token_accuracy": 0.19797009825706482,
      "num_tokens": 9765635.0,
      "step": 4265
    },
    {
      "entropy": 5.564598083496094,
      "epoch": 0.4101825168107589,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988844479710369,
      "loss": 5.4176,
      "mean_token_accuracy": 0.1858292818069458,
      "num_tokens": 9777512.0,
      "step": 4270
    },
    {
      "entropy": 5.581813335418701,
      "epoch": 0.4106628242074928,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004988810356770884,
      "loss": 5.4259,
      "mean_token_accuracy": 0.182652647793293,
      "num_tokens": 9790128.0,
      "step": 4275
    },
    {
      "entropy": 5.5544164180755615,
      "epoch": 0.4111431316042267,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498877618185287,
      "loss": 5.3703,
      "mean_token_accuracy": 0.19317726492881776,
      "num_tokens": 9802549.0,
      "step": 4280
    },
    {
      "entropy": 5.475317859649659,
      "epoch": 0.4116234390009606,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988741954957121,
      "loss": 5.3106,
      "mean_token_accuracy": 0.19570045918226242,
      "num_tokens": 9813736.0,
      "step": 4285
    },
    {
      "entropy": 5.428925800323486,
      "epoch": 0.41210374639769454,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988707676084432,
      "loss": 5.3157,
      "mean_token_accuracy": 0.19985667020082473,
      "num_tokens": 9823785.0,
      "step": 4290
    },
    {
      "entropy": 5.524678373336792,
      "epoch": 0.4125840537944284,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988673345235597,
      "loss": 5.269,
      "mean_token_accuracy": 0.1980641022324562,
      "num_tokens": 9834910.0,
      "step": 4295
    },
    {
      "entropy": 5.533667707443238,
      "epoch": 0.41306436119116235,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988638962411416,
      "loss": 5.3092,
      "mean_token_accuracy": 0.1934407815337181,
      "num_tokens": 9845593.0,
      "step": 4300
    },
    {
      "entropy": 5.438078308105469,
      "epoch": 0.41354466858789624,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004988604527612685,
      "loss": 5.2094,
      "mean_token_accuracy": 0.207307867705822,
      "num_tokens": 9856763.0,
      "step": 4305
    },
    {
      "entropy": 5.557668399810791,
      "epoch": 0.4140249759846302,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988570040840205,
      "loss": 5.4396,
      "mean_token_accuracy": 0.18502080589532852,
      "num_tokens": 9869528.0,
      "step": 4310
    },
    {
      "entropy": 5.517323637008667,
      "epoch": 0.41450528338136405,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988535502094774,
      "loss": 5.3325,
      "mean_token_accuracy": 0.19549246728420258,
      "num_tokens": 9881170.0,
      "step": 4315
    },
    {
      "entropy": 5.527880764007568,
      "epoch": 0.414985590778098,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988500911377198,
      "loss": 5.4255,
      "mean_token_accuracy": 0.18764246851205826,
      "num_tokens": 9893119.0,
      "step": 4320
    },
    {
      "entropy": 5.423438262939453,
      "epoch": 0.41546589817483187,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004988466268688276,
      "loss": 5.2627,
      "mean_token_accuracy": 0.20306207984685898,
      "num_tokens": 9905339.0,
      "step": 4325
    },
    {
      "entropy": 5.555384874343872,
      "epoch": 0.4159462055715658,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004988431574028814,
      "loss": 5.3633,
      "mean_token_accuracy": 0.19620103389024734,
      "num_tokens": 9917500.0,
      "step": 4330
    },
    {
      "entropy": 5.523201417922974,
      "epoch": 0.4164265129682997,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004988396827399618,
      "loss": 5.4231,
      "mean_token_accuracy": 0.19124320894479752,
      "num_tokens": 9929667.0,
      "step": 4335
    },
    {
      "entropy": 5.538033056259155,
      "epoch": 0.4169068203650336,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988362028801495,
      "loss": 5.3526,
      "mean_token_accuracy": 0.19107616543769837,
      "num_tokens": 9941102.0,
      "step": 4340
    },
    {
      "entropy": 5.46150393486023,
      "epoch": 0.4173871277617675,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988327178235253,
      "loss": 5.2557,
      "mean_token_accuracy": 0.20461782515048982,
      "num_tokens": 9951986.0,
      "step": 4345
    },
    {
      "entropy": 5.464718580245972,
      "epoch": 0.41786743515850144,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004988292275701699,
      "loss": 5.2626,
      "mean_token_accuracy": 0.2002312421798706,
      "num_tokens": 9964486.0,
      "step": 4350
    },
    {
      "entropy": 5.49985671043396,
      "epoch": 0.4183477425552353,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988257321201646,
      "loss": 5.3613,
      "mean_token_accuracy": 0.1955868422985077,
      "num_tokens": 9975909.0,
      "step": 4355
    },
    {
      "entropy": 5.55556583404541,
      "epoch": 0.41882804995196926,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988222314735902,
      "loss": 5.3655,
      "mean_token_accuracy": 0.19315043240785598,
      "num_tokens": 9986951.0,
      "step": 4360
    },
    {
      "entropy": 5.574152660369873,
      "epoch": 0.41930835734870314,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988187256305284,
      "loss": 5.4691,
      "mean_token_accuracy": 0.18375526517629623,
      "num_tokens": 9999234.0,
      "step": 4365
    },
    {
      "entropy": 5.496746015548706,
      "epoch": 0.4197886647454371,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988152145910603,
      "loss": 5.3249,
      "mean_token_accuracy": 0.19960517138242723,
      "num_tokens": 10010178.0,
      "step": 4370
    },
    {
      "entropy": 5.469662141799927,
      "epoch": 0.420268972142171,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988116983552675,
      "loss": 5.2854,
      "mean_token_accuracy": 0.19531275182962418,
      "num_tokens": 10021183.0,
      "step": 4375
    },
    {
      "entropy": 5.641615295410157,
      "epoch": 0.4207492795389049,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988081769232317,
      "loss": 5.5787,
      "mean_token_accuracy": 0.17796630561351776,
      "num_tokens": 10033686.0,
      "step": 4380
    },
    {
      "entropy": 5.461948680877685,
      "epoch": 0.42122958693563883,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988046502950346,
      "loss": 5.2841,
      "mean_token_accuracy": 0.19567082226276397,
      "num_tokens": 10045923.0,
      "step": 4385
    },
    {
      "entropy": 5.335393571853638,
      "epoch": 0.4217098943323727,
      "grad_norm": 0.859375,
      "learning_rate": 0.000498801118470758,
      "loss": 5.1912,
      "mean_token_accuracy": 0.20154503285884856,
      "num_tokens": 10057196.0,
      "step": 4390
    },
    {
      "entropy": 5.4671718120574955,
      "epoch": 0.42219020172910665,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498797581450484,
      "loss": 5.3699,
      "mean_token_accuracy": 0.1907932788133621,
      "num_tokens": 10069655.0,
      "step": 4395
    },
    {
      "entropy": 5.550377655029297,
      "epoch": 0.42267050912584053,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004987940392342948,
      "loss": 5.2483,
      "mean_token_accuracy": 0.1976744294166565,
      "num_tokens": 10080876.0,
      "step": 4400
    },
    {
      "entropy": 5.534447097778321,
      "epoch": 0.42315081652257447,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004987904918222726,
      "loss": 5.3552,
      "mean_token_accuracy": 0.19047154486179352,
      "num_tokens": 10091986.0,
      "step": 4405
    },
    {
      "entropy": 5.534992027282715,
      "epoch": 0.42363112391930835,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987869392144996,
      "loss": 5.4398,
      "mean_token_accuracy": 0.19162497520446778,
      "num_tokens": 10104027.0,
      "step": 4410
    },
    {
      "entropy": 5.52318263053894,
      "epoch": 0.4241114313160423,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987833814110584,
      "loss": 5.3064,
      "mean_token_accuracy": 0.19698891639709473,
      "num_tokens": 10114665.0,
      "step": 4415
    },
    {
      "entropy": 5.497555065155029,
      "epoch": 0.42459173871277617,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987798184120316,
      "loss": 5.3206,
      "mean_token_accuracy": 0.19676625728607178,
      "num_tokens": 10126032.0,
      "step": 4420
    },
    {
      "entropy": 5.534703922271729,
      "epoch": 0.4250720461095101,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987762502175018,
      "loss": 5.3858,
      "mean_token_accuracy": 0.19109898060560226,
      "num_tokens": 10137256.0,
      "step": 4425
    },
    {
      "entropy": 5.452436637878418,
      "epoch": 0.425552353506244,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498772676827552,
      "loss": 5.2682,
      "mean_token_accuracy": 0.19692862033843994,
      "num_tokens": 10149445.0,
      "step": 4430
    },
    {
      "entropy": 5.571733474731445,
      "epoch": 0.4260326609029779,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987690982422652,
      "loss": 5.4037,
      "mean_token_accuracy": 0.18802779018878937,
      "num_tokens": 10161607.0,
      "step": 4435
    },
    {
      "entropy": 5.538624715805054,
      "epoch": 0.4265129682997118,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987655144617243,
      "loss": 5.4126,
      "mean_token_accuracy": 0.19259897768497466,
      "num_tokens": 10173184.0,
      "step": 4440
    },
    {
      "entropy": 5.47358283996582,
      "epoch": 0.42699327569644574,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987619254860126,
      "loss": 5.2774,
      "mean_token_accuracy": 0.19967445135116577,
      "num_tokens": 10184617.0,
      "step": 4445
    },
    {
      "entropy": 5.516195583343506,
      "epoch": 0.4274735830931796,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987583313152134,
      "loss": 5.3133,
      "mean_token_accuracy": 0.19398918002843857,
      "num_tokens": 10195608.0,
      "step": 4450
    },
    {
      "entropy": 5.499154186248779,
      "epoch": 0.42795389048991356,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987547319494104,
      "loss": 5.3918,
      "mean_token_accuracy": 0.19446780085563659,
      "num_tokens": 10206763.0,
      "step": 4455
    },
    {
      "entropy": 5.52622652053833,
      "epoch": 0.42843419788664744,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004987511273886867,
      "loss": 5.3461,
      "mean_token_accuracy": 0.19333059340715408,
      "num_tokens": 10218714.0,
      "step": 4460
    },
    {
      "entropy": 5.5156656265258786,
      "epoch": 0.4289145052833814,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987475176331263,
      "loss": 5.3669,
      "mean_token_accuracy": 0.1908699557185173,
      "num_tokens": 10229902.0,
      "step": 4465
    },
    {
      "entropy": 5.496946096420288,
      "epoch": 0.42939481268011526,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004987439026828129,
      "loss": 5.2409,
      "mean_token_accuracy": 0.19724130481481553,
      "num_tokens": 10241578.0,
      "step": 4470
    },
    {
      "entropy": 5.42222957611084,
      "epoch": 0.4298751200768492,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987402825378305,
      "loss": 5.2035,
      "mean_token_accuracy": 0.20090758353471755,
      "num_tokens": 10252109.0,
      "step": 4475
    },
    {
      "entropy": 5.5068761825561525,
      "epoch": 0.4303554274735831,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004987366571982631,
      "loss": 5.38,
      "mean_token_accuracy": 0.1923118144273758,
      "num_tokens": 10263357.0,
      "step": 4480
    },
    {
      "entropy": 5.550309705734253,
      "epoch": 0.430835734870317,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987330266641948,
      "loss": 5.3643,
      "mean_token_accuracy": 0.19260043650865555,
      "num_tokens": 10275536.0,
      "step": 4485
    },
    {
      "entropy": 5.575230693817138,
      "epoch": 0.4313160422670509,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987293909357101,
      "loss": 5.3461,
      "mean_token_accuracy": 0.20100529342889786,
      "num_tokens": 10286901.0,
      "step": 4490
    },
    {
      "entropy": 5.437625217437744,
      "epoch": 0.43179634966378483,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987257500128933,
      "loss": 5.2572,
      "mean_token_accuracy": 0.1957649528980255,
      "num_tokens": 10298961.0,
      "step": 4495
    },
    {
      "entropy": 5.524259567260742,
      "epoch": 0.4322766570605187,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987221038958288,
      "loss": 5.403,
      "mean_token_accuracy": 0.19641273021697997,
      "num_tokens": 10310911.0,
      "step": 4500
    },
    {
      "entropy": 5.567202568054199,
      "epoch": 0.43275696445725265,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987184525846015,
      "loss": 5.3775,
      "mean_token_accuracy": 0.1914454460144043,
      "num_tokens": 10322267.0,
      "step": 4505
    },
    {
      "entropy": 5.494447660446167,
      "epoch": 0.4332372718539865,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987147960792958,
      "loss": 5.4018,
      "mean_token_accuracy": 0.1945730999112129,
      "num_tokens": 10335111.0,
      "step": 4510
    },
    {
      "entropy": 5.585152101516724,
      "epoch": 0.43371757925072046,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987111343799971,
      "loss": 5.3466,
      "mean_token_accuracy": 0.19717937111854553,
      "num_tokens": 10345672.0,
      "step": 4515
    },
    {
      "entropy": 5.579340267181396,
      "epoch": 0.43419788664745435,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00049870746748679,
      "loss": 5.3364,
      "mean_token_accuracy": 0.19359557032585145,
      "num_tokens": 10357369.0,
      "step": 4520
    },
    {
      "entropy": 5.46253399848938,
      "epoch": 0.4346781940441883,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987037953997598,
      "loss": 5.3542,
      "mean_token_accuracy": 0.19162115156650544,
      "num_tokens": 10368842.0,
      "step": 4525
    },
    {
      "entropy": 5.506046867370605,
      "epoch": 0.43515850144092216,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987001181189918,
      "loss": 5.3019,
      "mean_token_accuracy": 0.19462748765945434,
      "num_tokens": 10380096.0,
      "step": 4530
    },
    {
      "entropy": 5.433768033981323,
      "epoch": 0.4356388088376561,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986964356445713,
      "loss": 5.3279,
      "mean_token_accuracy": 0.19543916881084442,
      "num_tokens": 10391996.0,
      "step": 4535
    },
    {
      "entropy": 5.506909990310669,
      "epoch": 0.43611911623439004,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986927479765837,
      "loss": 5.2586,
      "mean_token_accuracy": 0.1950858324766159,
      "num_tokens": 10403607.0,
      "step": 4540
    },
    {
      "entropy": 5.500537729263305,
      "epoch": 0.4365994236311239,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986890551151148,
      "loss": 5.3027,
      "mean_token_accuracy": 0.19058295637369155,
      "num_tokens": 10413580.0,
      "step": 4545
    },
    {
      "entropy": 5.453525447845459,
      "epoch": 0.43707973102785785,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986853570602503,
      "loss": 5.3248,
      "mean_token_accuracy": 0.19454682171344756,
      "num_tokens": 10426456.0,
      "step": 4550
    },
    {
      "entropy": 5.570721864700317,
      "epoch": 0.43756003842459174,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004986816538120758,
      "loss": 5.3464,
      "mean_token_accuracy": 0.18605243414640427,
      "num_tokens": 10438869.0,
      "step": 4555
    },
    {
      "entropy": 5.467431879043579,
      "epoch": 0.43804034582132567,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986779453706778,
      "loss": 5.3617,
      "mean_token_accuracy": 0.18677616268396377,
      "num_tokens": 10450672.0,
      "step": 4560
    },
    {
      "entropy": 5.506100177764893,
      "epoch": 0.43852065321805955,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986742317361419,
      "loss": 5.2719,
      "mean_token_accuracy": 0.19866655319929122,
      "num_tokens": 10461890.0,
      "step": 4565
    },
    {
      "entropy": 5.571147489547729,
      "epoch": 0.4390009606147935,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986705129085546,
      "loss": 5.4027,
      "mean_token_accuracy": 0.18525223433971405,
      "num_tokens": 10473866.0,
      "step": 4570
    },
    {
      "entropy": 5.531253671646118,
      "epoch": 0.43948126801152737,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986667888880021,
      "loss": 5.3279,
      "mean_token_accuracy": 0.19253301173448562,
      "num_tokens": 10484889.0,
      "step": 4575
    },
    {
      "entropy": 5.478489875793457,
      "epoch": 0.4399615754082613,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986630596745709,
      "loss": 5.366,
      "mean_token_accuracy": 0.19603030383586884,
      "num_tokens": 10496108.0,
      "step": 4580
    },
    {
      "entropy": 5.481814002990722,
      "epoch": 0.4404418828049952,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986593252683477,
      "loss": 5.3077,
      "mean_token_accuracy": 0.19383252114057542,
      "num_tokens": 10505472.0,
      "step": 4585
    },
    {
      "entropy": 5.359921264648437,
      "epoch": 0.4409221902017291,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986555856694191,
      "loss": 5.2242,
      "mean_token_accuracy": 0.19945572316646576,
      "num_tokens": 10516954.0,
      "step": 4590
    },
    {
      "entropy": 5.584690189361572,
      "epoch": 0.441402497598463,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986518408778718,
      "loss": 5.341,
      "mean_token_accuracy": 0.19385933727025986,
      "num_tokens": 10528166.0,
      "step": 4595
    },
    {
      "entropy": 5.464456033706665,
      "epoch": 0.44188280499519694,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986480908937929,
      "loss": 5.2583,
      "mean_token_accuracy": 0.198069229722023,
      "num_tokens": 10538112.0,
      "step": 4600
    },
    {
      "entropy": 5.505298757553101,
      "epoch": 0.4423631123919308,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986443357172695,
      "loss": 5.4021,
      "mean_token_accuracy": 0.19536336809396743,
      "num_tokens": 10549888.0,
      "step": 4605
    },
    {
      "entropy": 5.657861185073853,
      "epoch": 0.44284341978866476,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004986405753483887,
      "loss": 5.4699,
      "mean_token_accuracy": 0.1838005244731903,
      "num_tokens": 10561710.0,
      "step": 4610
    },
    {
      "entropy": 5.473537921905518,
      "epoch": 0.44332372718539864,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004986368097872377,
      "loss": 5.3241,
      "mean_token_accuracy": 0.19243624061346054,
      "num_tokens": 10574564.0,
      "step": 4615
    },
    {
      "entropy": 5.540151643753052,
      "epoch": 0.4438040345821326,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986330390339042,
      "loss": 5.3188,
      "mean_token_accuracy": 0.19761135280132294,
      "num_tokens": 10586639.0,
      "step": 4620
    },
    {
      "entropy": 5.506497859954834,
      "epoch": 0.44428434197886646,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986292630884755,
      "loss": 5.3144,
      "mean_token_accuracy": 0.1980729505419731,
      "num_tokens": 10598730.0,
      "step": 4625
    },
    {
      "entropy": 5.433584356307984,
      "epoch": 0.4447646493756004,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986254819510393,
      "loss": 5.2295,
      "mean_token_accuracy": 0.2066010907292366,
      "num_tokens": 10610352.0,
      "step": 4630
    },
    {
      "entropy": 5.467961835861206,
      "epoch": 0.4452449567723343,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986216956216835,
      "loss": 5.3149,
      "mean_token_accuracy": 0.19723764508962632,
      "num_tokens": 10621951.0,
      "step": 4635
    },
    {
      "entropy": 5.48427186012268,
      "epoch": 0.4457252641690682,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498617904100496,
      "loss": 5.269,
      "mean_token_accuracy": 0.19730617105960846,
      "num_tokens": 10633207.0,
      "step": 4640
    },
    {
      "entropy": 5.5214195728302,
      "epoch": 0.4462055715658021,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986141073875646,
      "loss": 5.3485,
      "mean_token_accuracy": 0.1888623684644699,
      "num_tokens": 10645853.0,
      "step": 4645
    },
    {
      "entropy": 5.466204071044922,
      "epoch": 0.44668587896253603,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986103054829779,
      "loss": 5.2731,
      "mean_token_accuracy": 0.197029510140419,
      "num_tokens": 10656892.0,
      "step": 4650
    },
    {
      "entropy": 5.430473184585571,
      "epoch": 0.4471661863592699,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004986064983868237,
      "loss": 5.2464,
      "mean_token_accuracy": 0.19406193494796753,
      "num_tokens": 10670110.0,
      "step": 4655
    },
    {
      "entropy": 5.522889232635498,
      "epoch": 0.44764649375600385,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986026860991906,
      "loss": 5.3824,
      "mean_token_accuracy": 0.19317276924848556,
      "num_tokens": 10681255.0,
      "step": 4660
    },
    {
      "entropy": 5.506927394866944,
      "epoch": 0.44812680115273773,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004985988686201672,
      "loss": 5.4456,
      "mean_token_accuracy": 0.19067862927913665,
      "num_tokens": 10692631.0,
      "step": 4665
    },
    {
      "entropy": 5.498530721664428,
      "epoch": 0.44860710854947167,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985950459498419,
      "loss": 5.2764,
      "mean_token_accuracy": 0.19727545231580734,
      "num_tokens": 10704880.0,
      "step": 4670
    },
    {
      "entropy": 5.481644058227539,
      "epoch": 0.44908741594620555,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985912180883037,
      "loss": 5.267,
      "mean_token_accuracy": 0.19885830134153365,
      "num_tokens": 10715561.0,
      "step": 4675
    },
    {
      "entropy": 5.432435321807861,
      "epoch": 0.4495677233429395,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985873850356411,
      "loss": 5.2767,
      "mean_token_accuracy": 0.19294027239084244,
      "num_tokens": 10727232.0,
      "step": 4680
    },
    {
      "entropy": 5.450063991546631,
      "epoch": 0.45004803073967337,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985835467919436,
      "loss": 5.2946,
      "mean_token_accuracy": 0.1977899894118309,
      "num_tokens": 10739404.0,
      "step": 4685
    },
    {
      "entropy": 5.486606788635254,
      "epoch": 0.4505283381364073,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985797033572999,
      "loss": 5.3282,
      "mean_token_accuracy": 0.19158132523298263,
      "num_tokens": 10751948.0,
      "step": 4690
    },
    {
      "entropy": 5.4498594284057615,
      "epoch": 0.4510086455331412,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985758547317994,
      "loss": 5.2983,
      "mean_token_accuracy": 0.19393019676208495,
      "num_tokens": 10764611.0,
      "step": 4695
    },
    {
      "entropy": 5.533370399475098,
      "epoch": 0.4514889529298751,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985720009155315,
      "loss": 5.3189,
      "mean_token_accuracy": 0.19145917296409606,
      "num_tokens": 10775954.0,
      "step": 4700
    },
    {
      "entropy": 5.4836106300354,
      "epoch": 0.45196926032660906,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004985681419085856,
      "loss": 5.3281,
      "mean_token_accuracy": 0.18578007370233535,
      "num_tokens": 10788723.0,
      "step": 4705
    },
    {
      "entropy": 5.504494524002075,
      "epoch": 0.45244956772334294,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985642777110513,
      "loss": 5.3192,
      "mean_token_accuracy": 0.19870994240045547,
      "num_tokens": 10799879.0,
      "step": 4710
    },
    {
      "entropy": 5.444184827804565,
      "epoch": 0.4529298751200769,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004985604083230183,
      "loss": 5.278,
      "mean_token_accuracy": 0.19549098014831542,
      "num_tokens": 10811838.0,
      "step": 4715
    },
    {
      "entropy": 5.476307773590088,
      "epoch": 0.45341018251681076,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004985565337445765,
      "loss": 5.2976,
      "mean_token_accuracy": 0.19534050822257995,
      "num_tokens": 10822910.0,
      "step": 4720
    },
    {
      "entropy": 5.545818233489991,
      "epoch": 0.4538904899135447,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004985526539758158,
      "loss": 5.3536,
      "mean_token_accuracy": 0.19093685001134872,
      "num_tokens": 10835344.0,
      "step": 4725
    },
    {
      "entropy": 5.473781394958496,
      "epoch": 0.4543707973102786,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985487690168263,
      "loss": 5.3433,
      "mean_token_accuracy": 0.19793201237916946,
      "num_tokens": 10846043.0,
      "step": 4730
    },
    {
      "entropy": 5.483222532272339,
      "epoch": 0.4548511047070125,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498544878867698,
      "loss": 5.2471,
      "mean_token_accuracy": 0.20124554485082627,
      "num_tokens": 10857783.0,
      "step": 4735
    },
    {
      "entropy": 5.502796602249146,
      "epoch": 0.4553314121037464,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985409835285215,
      "loss": 5.3284,
      "mean_token_accuracy": 0.19286340475082397,
      "num_tokens": 10870527.0,
      "step": 4740
    },
    {
      "entropy": 5.451869821548462,
      "epoch": 0.45581171950048033,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985370829993873,
      "loss": 5.3101,
      "mean_token_accuracy": 0.1995761588215828,
      "num_tokens": 10882285.0,
      "step": 4745
    },
    {
      "entropy": 5.519151926040649,
      "epoch": 0.4562920268972142,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985331772803857,
      "loss": 5.3397,
      "mean_token_accuracy": 0.19749111980199813,
      "num_tokens": 10895319.0,
      "step": 4750
    },
    {
      "entropy": 5.546896362304688,
      "epoch": 0.45677233429394815,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985292663716074,
      "loss": 5.3178,
      "mean_token_accuracy": 0.1997792139649391,
      "num_tokens": 10906253.0,
      "step": 4755
    },
    {
      "entropy": 5.323689603805542,
      "epoch": 0.457252641690682,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985253502731435,
      "loss": 5.2064,
      "mean_token_accuracy": 0.20633554905653,
      "num_tokens": 10918197.0,
      "step": 4760
    },
    {
      "entropy": 5.48868842124939,
      "epoch": 0.45773294908741596,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004985214289850845,
      "loss": 5.4051,
      "mean_token_accuracy": 0.18659979701042176,
      "num_tokens": 10930771.0,
      "step": 4765
    },
    {
      "entropy": 5.524543571472168,
      "epoch": 0.45821325648414984,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004985175025075217,
      "loss": 5.2965,
      "mean_token_accuracy": 0.19515040069818496,
      "num_tokens": 10942759.0,
      "step": 4770
    },
    {
      "entropy": 5.671936845779419,
      "epoch": 0.4586935638808838,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985135708405462,
      "loss": 5.5023,
      "mean_token_accuracy": 0.1835600718855858,
      "num_tokens": 10953557.0,
      "step": 4775
    },
    {
      "entropy": 5.4718035697937015,
      "epoch": 0.45917387127761766,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985096339842493,
      "loss": 5.2753,
      "mean_token_accuracy": 0.20337168574333192,
      "num_tokens": 10963142.0,
      "step": 4780
    },
    {
      "entropy": 5.392521715164184,
      "epoch": 0.4596541786743516,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985056919387224,
      "loss": 5.2352,
      "mean_token_accuracy": 0.2007603958249092,
      "num_tokens": 10974321.0,
      "step": 4785
    },
    {
      "entropy": 5.5639253616333,
      "epoch": 0.4601344860710855,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985017447040569,
      "loss": 5.439,
      "mean_token_accuracy": 0.1905608966946602,
      "num_tokens": 10985524.0,
      "step": 4790
    },
    {
      "entropy": 5.547641324996948,
      "epoch": 0.4606147934678194,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984977922803447,
      "loss": 5.3173,
      "mean_token_accuracy": 0.19772396385669708,
      "num_tokens": 10997606.0,
      "step": 4795
    },
    {
      "entropy": 5.467640161514282,
      "epoch": 0.4610951008645533,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004984938346676772,
      "loss": 5.3181,
      "mean_token_accuracy": 0.1880475327372551,
      "num_tokens": 11010692.0,
      "step": 4800
    },
    {
      "entropy": 5.430249834060669,
      "epoch": 0.46157540826128723,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004984898718661468,
      "loss": 5.2577,
      "mean_token_accuracy": 0.2035054549574852,
      "num_tokens": 11022517.0,
      "step": 4805
    },
    {
      "entropy": 5.480600738525391,
      "epoch": 0.4620557156580211,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004984859038758451,
      "loss": 5.2538,
      "mean_token_accuracy": 0.20069904923439025,
      "num_tokens": 11033141.0,
      "step": 4810
    },
    {
      "entropy": 5.440704441070556,
      "epoch": 0.46253602305475505,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984819306968642,
      "loss": 5.2587,
      "mean_token_accuracy": 0.1992591381072998,
      "num_tokens": 11044619.0,
      "step": 4815
    },
    {
      "entropy": 5.529746055603027,
      "epoch": 0.46301633045148893,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004984779523292966,
      "loss": 5.3095,
      "mean_token_accuracy": 0.1996643826365471,
      "num_tokens": 11055934.0,
      "step": 4820
    },
    {
      "entropy": 5.456304121017456,
      "epoch": 0.46349663784822287,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984739687732345,
      "loss": 5.1932,
      "mean_token_accuracy": 0.20430234223604202,
      "num_tokens": 11066203.0,
      "step": 4825
    },
    {
      "entropy": 5.294573020935059,
      "epoch": 0.46397694524495675,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004984699800287705,
      "loss": 5.1439,
      "mean_token_accuracy": 0.20539496690034867,
      "num_tokens": 11079664.0,
      "step": 4830
    },
    {
      "entropy": 5.391605186462402,
      "epoch": 0.4644572526416907,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000498465986095997,
      "loss": 5.2028,
      "mean_token_accuracy": 0.20457594841718674,
      "num_tokens": 11091186.0,
      "step": 4835
    },
    {
      "entropy": 5.495221424102783,
      "epoch": 0.46493756003842457,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984619869750069,
      "loss": 5.3313,
      "mean_token_accuracy": 0.19135818332433702,
      "num_tokens": 11102710.0,
      "step": 4840
    },
    {
      "entropy": 5.393061113357544,
      "epoch": 0.4654178674351585,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498457982665893,
      "loss": 5.2222,
      "mean_token_accuracy": 0.20273594111204146,
      "num_tokens": 11114746.0,
      "step": 4845
    },
    {
      "entropy": 5.48639497756958,
      "epoch": 0.4658981748318924,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984539731687483,
      "loss": 5.2975,
      "mean_token_accuracy": 0.19519798308610917,
      "num_tokens": 11126572.0,
      "step": 4850
    },
    {
      "entropy": 5.427862167358398,
      "epoch": 0.4663784822286263,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984499584836659,
      "loss": 5.1845,
      "mean_token_accuracy": 0.2038702443242073,
      "num_tokens": 11137830.0,
      "step": 4855
    },
    {
      "entropy": 5.402318859100342,
      "epoch": 0.4668587896253602,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498445938610739,
      "loss": 5.2247,
      "mean_token_accuracy": 0.1985754981637001,
      "num_tokens": 11148860.0,
      "step": 4860
    },
    {
      "entropy": 5.510347509384156,
      "epoch": 0.46733909702209414,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984419135500608,
      "loss": 5.3408,
      "mean_token_accuracy": 0.18952501863241195,
      "num_tokens": 11161311.0,
      "step": 4865
    },
    {
      "entropy": 5.4862135887146,
      "epoch": 0.4678194044188281,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004984378833017249,
      "loss": 5.2453,
      "mean_token_accuracy": 0.19657314866781234,
      "num_tokens": 11173124.0,
      "step": 4870
    },
    {
      "entropy": 5.405213022232056,
      "epoch": 0.46829971181556196,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984338478658248,
      "loss": 5.3077,
      "mean_token_accuracy": 0.1948263332247734,
      "num_tokens": 11184879.0,
      "step": 4875
    },
    {
      "entropy": 5.519789314270019,
      "epoch": 0.4687800192122959,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984298072424542,
      "loss": 5.3156,
      "mean_token_accuracy": 0.19586896896362305,
      "num_tokens": 11196243.0,
      "step": 4880
    },
    {
      "entropy": 5.473620462417602,
      "epoch": 0.4692603266090298,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498425761431707,
      "loss": 5.1839,
      "mean_token_accuracy": 0.2131851315498352,
      "num_tokens": 11207485.0,
      "step": 4885
    },
    {
      "entropy": 5.417775678634643,
      "epoch": 0.4697406340057637,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498421710433677,
      "loss": 5.2255,
      "mean_token_accuracy": 0.1948077842593193,
      "num_tokens": 11219891.0,
      "step": 4890
    },
    {
      "entropy": 5.54314079284668,
      "epoch": 0.4702209414024976,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004984176542484584,
      "loss": 5.3247,
      "mean_token_accuracy": 0.1956147536635399,
      "num_tokens": 11231329.0,
      "step": 4895
    },
    {
      "entropy": 5.435543823242187,
      "epoch": 0.47070124879923153,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984135928761452,
      "loss": 5.2117,
      "mean_token_accuracy": 0.20288991183042526,
      "num_tokens": 11241367.0,
      "step": 4900
    },
    {
      "entropy": 5.44714093208313,
      "epoch": 0.4711815561959654,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984095263168317,
      "loss": 5.2858,
      "mean_token_accuracy": 0.19364715069532396,
      "num_tokens": 11254532.0,
      "step": 4905
    },
    {
      "entropy": 5.544944953918457,
      "epoch": 0.47166186359269935,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004984054545706124,
      "loss": 5.3768,
      "mean_token_accuracy": 0.1902521938085556,
      "num_tokens": 11265223.0,
      "step": 4910
    },
    {
      "entropy": 5.443441390991211,
      "epoch": 0.47214217098943323,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000498401377637582,
      "loss": 5.3013,
      "mean_token_accuracy": 0.19836317002773285,
      "num_tokens": 11278228.0,
      "step": 4915
    },
    {
      "entropy": 5.577527093887329,
      "epoch": 0.47262247838616717,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000498397295517835,
      "loss": 5.4154,
      "mean_token_accuracy": 0.1912369191646576,
      "num_tokens": 11289654.0,
      "step": 4920
    },
    {
      "entropy": 5.449563312530517,
      "epoch": 0.47310278578290105,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004983932082114659,
      "loss": 5.1923,
      "mean_token_accuracy": 0.20263250917196274,
      "num_tokens": 11301911.0,
      "step": 4925
    },
    {
      "entropy": 5.550692844390869,
      "epoch": 0.473583093179635,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983891157185699,
      "loss": 5.3589,
      "mean_token_accuracy": 0.19057601392269136,
      "num_tokens": 11312945.0,
      "step": 4930
    },
    {
      "entropy": 5.617296600341797,
      "epoch": 0.47406340057636887,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983850180392421,
      "loss": 5.4156,
      "mean_token_accuracy": 0.184731362760067,
      "num_tokens": 11324126.0,
      "step": 4935
    },
    {
      "entropy": 5.5141273021698,
      "epoch": 0.4745437079731028,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983809151735775,
      "loss": 5.3503,
      "mean_token_accuracy": 0.1906953737139702,
      "num_tokens": 11336395.0,
      "step": 4940
    },
    {
      "entropy": 5.499377822875976,
      "epoch": 0.4750240153698367,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004983768071216713,
      "loss": 5.2561,
      "mean_token_accuracy": 0.19905748665332795,
      "num_tokens": 11347387.0,
      "step": 4945
    },
    {
      "entropy": 5.396799373626709,
      "epoch": 0.4755043227665706,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983726938836189,
      "loss": 5.2422,
      "mean_token_accuracy": 0.20498072654008864,
      "num_tokens": 11358467.0,
      "step": 4950
    },
    {
      "entropy": 5.49179253578186,
      "epoch": 0.4759846301633045,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983685754595159,
      "loss": 5.4096,
      "mean_token_accuracy": 0.19101459980010987,
      "num_tokens": 11370322.0,
      "step": 4955
    },
    {
      "entropy": 5.400127029418945,
      "epoch": 0.47646493756003844,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983644518494578,
      "loss": 5.2078,
      "mean_token_accuracy": 0.20537745207548141,
      "num_tokens": 11381719.0,
      "step": 4960
    },
    {
      "entropy": 5.4417760372161865,
      "epoch": 0.4769452449567723,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983603230535403,
      "loss": 5.2271,
      "mean_token_accuracy": 0.20112286359071732,
      "num_tokens": 11393561.0,
      "step": 4965
    },
    {
      "entropy": 5.495945167541504,
      "epoch": 0.47742555235350626,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004983561890718594,
      "loss": 5.3273,
      "mean_token_accuracy": 0.1953832298517227,
      "num_tokens": 11405411.0,
      "step": 4970
    },
    {
      "entropy": 5.537110948562622,
      "epoch": 0.47790585975024014,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498352049904511,
      "loss": 5.4195,
      "mean_token_accuracy": 0.1894658625125885,
      "num_tokens": 11417419.0,
      "step": 4975
    },
    {
      "entropy": 5.492123174667358,
      "epoch": 0.4783861671469741,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983479055515914,
      "loss": 5.2302,
      "mean_token_accuracy": 0.19503100365400314,
      "num_tokens": 11428145.0,
      "step": 4980
    },
    {
      "entropy": 5.382611560821533,
      "epoch": 0.47886647454370795,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004983437560131964,
      "loss": 5.1931,
      "mean_token_accuracy": 0.20153277218341828,
      "num_tokens": 11439224.0,
      "step": 4985
    },
    {
      "entropy": 5.459662389755249,
      "epoch": 0.4793467819404419,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983396012894228,
      "loss": 5.2877,
      "mean_token_accuracy": 0.19697522670030593,
      "num_tokens": 11451731.0,
      "step": 4990
    },
    {
      "entropy": 5.489697217941284,
      "epoch": 0.47982708933717577,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983354413803666,
      "loss": 5.2723,
      "mean_token_accuracy": 0.20329660624265672,
      "num_tokens": 11463058.0,
      "step": 4995
    },
    {
      "entropy": 5.602862405776977,
      "epoch": 0.4803073967339097,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983312762861248,
      "loss": 5.3766,
      "mean_token_accuracy": 0.19309359192848205,
      "num_tokens": 11472618.0,
      "step": 5000
    },
    {
      "entropy": 5.436448240280152,
      "epoch": 0.4807877041306436,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983271060067939,
      "loss": 5.2597,
      "mean_token_accuracy": 0.19574105590581894,
      "num_tokens": 11483114.0,
      "step": 5005
    },
    {
      "entropy": 5.403754663467407,
      "epoch": 0.4812680115273775,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983229305424707,
      "loss": 5.2011,
      "mean_token_accuracy": 0.19807563424110414,
      "num_tokens": 11494281.0,
      "step": 5010
    },
    {
      "entropy": 5.481242561340332,
      "epoch": 0.4817483189241114,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004983187498932522,
      "loss": 5.3145,
      "mean_token_accuracy": 0.19510846138000487,
      "num_tokens": 11505962.0,
      "step": 5015
    },
    {
      "entropy": 5.499330186843872,
      "epoch": 0.48222862632084534,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004983145640592354,
      "loss": 5.3657,
      "mean_token_accuracy": 0.18856551945209504,
      "num_tokens": 11517558.0,
      "step": 5020
    },
    {
      "entropy": 5.483879709243775,
      "epoch": 0.4827089337175792,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983103730405176,
      "loss": 5.3623,
      "mean_token_accuracy": 0.19289948493242265,
      "num_tokens": 11529184.0,
      "step": 5025
    },
    {
      "entropy": 5.484569931030274,
      "epoch": 0.48318924111431316,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498306176837196,
      "loss": 5.2893,
      "mean_token_accuracy": 0.18928477317094802,
      "num_tokens": 11540727.0,
      "step": 5030
    },
    {
      "entropy": 5.440706968307495,
      "epoch": 0.48366954851104704,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983019754493681,
      "loss": 5.222,
      "mean_token_accuracy": 0.19865260273218155,
      "num_tokens": 11551510.0,
      "step": 5035
    },
    {
      "entropy": 5.502947568893433,
      "epoch": 0.484149855907781,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982977688771314,
      "loss": 5.353,
      "mean_token_accuracy": 0.19847064167261125,
      "num_tokens": 11563203.0,
      "step": 5040
    },
    {
      "entropy": 5.40673451423645,
      "epoch": 0.4846301633045149,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004982935571205835,
      "loss": 5.215,
      "mean_token_accuracy": 0.2036955922842026,
      "num_tokens": 11576013.0,
      "step": 5045
    },
    {
      "entropy": 5.325626945495605,
      "epoch": 0.4851104707012488,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004982893401798223,
      "loss": 5.1837,
      "mean_token_accuracy": 0.21459532976150514,
      "num_tokens": 11587535.0,
      "step": 5050
    },
    {
      "entropy": 5.416988277435303,
      "epoch": 0.48559077809798273,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004982851180549456,
      "loss": 5.2204,
      "mean_token_accuracy": 0.20377931892871856,
      "num_tokens": 11598487.0,
      "step": 5055
    },
    {
      "entropy": 5.380494689941406,
      "epoch": 0.4860710854947166,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982808907460515,
      "loss": 5.1061,
      "mean_token_accuracy": 0.21789742708206178,
      "num_tokens": 11609457.0,
      "step": 5060
    },
    {
      "entropy": 5.379251003265381,
      "epoch": 0.48655139289145055,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982766582532382,
      "loss": 5.1703,
      "mean_token_accuracy": 0.206756691634655,
      "num_tokens": 11620251.0,
      "step": 5065
    },
    {
      "entropy": 5.387517213821411,
      "epoch": 0.48703170028818443,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982724205766038,
      "loss": 5.1746,
      "mean_token_accuracy": 0.20464167147874832,
      "num_tokens": 11630956.0,
      "step": 5070
    },
    {
      "entropy": 5.436505603790283,
      "epoch": 0.48751200768491837,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004982681777162468,
      "loss": 5.2236,
      "mean_token_accuracy": 0.19775371849536896,
      "num_tokens": 11642560.0,
      "step": 5075
    },
    {
      "entropy": 5.353562116622925,
      "epoch": 0.48799231508165225,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982639296722657,
      "loss": 5.1914,
      "mean_token_accuracy": 0.2018374592065811,
      "num_tokens": 11654050.0,
      "step": 5080
    },
    {
      "entropy": 5.371949672698975,
      "epoch": 0.4884726224783862,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982596764447591,
      "loss": 5.3523,
      "mean_token_accuracy": 0.1960517406463623,
      "num_tokens": 11664947.0,
      "step": 5085
    },
    {
      "entropy": 5.506836891174316,
      "epoch": 0.48895292987512007,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982554180338258,
      "loss": 5.2498,
      "mean_token_accuracy": 0.19737753570079802,
      "num_tokens": 11676927.0,
      "step": 5090
    },
    {
      "entropy": 5.56851544380188,
      "epoch": 0.489433237271854,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004982511544395646,
      "loss": 5.3608,
      "mean_token_accuracy": 0.18899684697389602,
      "num_tokens": 11688573.0,
      "step": 5095
    },
    {
      "entropy": 5.375510406494141,
      "epoch": 0.4899135446685879,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982468856620745,
      "loss": 5.2578,
      "mean_token_accuracy": 0.19423613399267198,
      "num_tokens": 11698704.0,
      "step": 5100
    },
    {
      "entropy": 5.374054288864135,
      "epoch": 0.4903938520653218,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982426117014545,
      "loss": 5.1853,
      "mean_token_accuracy": 0.2054339423775673,
      "num_tokens": 11709466.0,
      "step": 5105
    },
    {
      "entropy": 5.459903144836426,
      "epoch": 0.4908741594620557,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982383325578041,
      "loss": 5.2851,
      "mean_token_accuracy": 0.19726966619491576,
      "num_tokens": 11721120.0,
      "step": 5110
    },
    {
      "entropy": 5.4925140857696535,
      "epoch": 0.49135446685878964,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982340482312226,
      "loss": 5.2708,
      "mean_token_accuracy": 0.1935970291495323,
      "num_tokens": 11732120.0,
      "step": 5115
    },
    {
      "entropy": 5.387935447692871,
      "epoch": 0.4918347742555235,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982297587218092,
      "loss": 5.1585,
      "mean_token_accuracy": 0.204207181930542,
      "num_tokens": 11743501.0,
      "step": 5120
    },
    {
      "entropy": 5.4283873558044435,
      "epoch": 0.49231508165225746,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004982254640296637,
      "loss": 5.2461,
      "mean_token_accuracy": 0.20177514851093292,
      "num_tokens": 11755051.0,
      "step": 5125
    },
    {
      "entropy": 5.521465635299682,
      "epoch": 0.49279538904899134,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004982211641548857,
      "loss": 5.3923,
      "mean_token_accuracy": 0.18937383741140365,
      "num_tokens": 11767663.0,
      "step": 5130
    },
    {
      "entropy": 5.450193262100219,
      "epoch": 0.4932756964457253,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004982168590975752,
      "loss": 5.256,
      "mean_token_accuracy": 0.20313976407051088,
      "num_tokens": 11778828.0,
      "step": 5135
    },
    {
      "entropy": 5.538891410827636,
      "epoch": 0.49375600384245916,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982125488578321,
      "loss": 5.4355,
      "mean_token_accuracy": 0.18950497061014177,
      "num_tokens": 11790654.0,
      "step": 5140
    },
    {
      "entropy": 5.489221286773682,
      "epoch": 0.4942363112391931,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982082334357563,
      "loss": 5.2177,
      "mean_token_accuracy": 0.19754690080881118,
      "num_tokens": 11801489.0,
      "step": 5145
    },
    {
      "entropy": 5.370436906814575,
      "epoch": 0.494716618635927,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982039128314481,
      "loss": 5.2229,
      "mean_token_accuracy": 0.2016189604997635,
      "num_tokens": 11813818.0,
      "step": 5150
    },
    {
      "entropy": 5.386388111114502,
      "epoch": 0.4951969260326609,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004981995870450079,
      "loss": 5.2316,
      "mean_token_accuracy": 0.19863317310810089,
      "num_tokens": 11824814.0,
      "step": 5155
    },
    {
      "entropy": 5.500951671600342,
      "epoch": 0.4956772334293948,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981952560765361,
      "loss": 5.2708,
      "mean_token_accuracy": 0.19523157477378844,
      "num_tokens": 11836252.0,
      "step": 5160
    },
    {
      "entropy": 5.4199676513671875,
      "epoch": 0.49615754082612873,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981909199261331,
      "loss": 5.2075,
      "mean_token_accuracy": 0.19706585705280305,
      "num_tokens": 11847715.0,
      "step": 5165
    },
    {
      "entropy": 5.309402847290039,
      "epoch": 0.4966378482228626,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004981865785938998,
      "loss": 5.2058,
      "mean_token_accuracy": 0.20216697901487352,
      "num_tokens": 11860309.0,
      "step": 5170
    },
    {
      "entropy": 5.3990058422088625,
      "epoch": 0.49711815561959655,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004981822320799367,
      "loss": 5.1964,
      "mean_token_accuracy": 0.20591772198677064,
      "num_tokens": 11872569.0,
      "step": 5175
    },
    {
      "entropy": 5.411104297637939,
      "epoch": 0.49759846301633043,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004981778803843449,
      "loss": 5.1837,
      "mean_token_accuracy": 0.2038358524441719,
      "num_tokens": 11884778.0,
      "step": 5180
    },
    {
      "entropy": 5.4573859691619875,
      "epoch": 0.49807877041306436,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981735235072256,
      "loss": 5.279,
      "mean_token_accuracy": 0.20060150176286698,
      "num_tokens": 11897324.0,
      "step": 5185
    },
    {
      "entropy": 5.510555553436279,
      "epoch": 0.49855907780979825,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981691614486796,
      "loss": 5.297,
      "mean_token_accuracy": 0.19576158821582795,
      "num_tokens": 11909145.0,
      "step": 5190
    },
    {
      "entropy": 5.4124188899993895,
      "epoch": 0.4990393852065322,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981647942088084,
      "loss": 5.1992,
      "mean_token_accuracy": 0.20338936448097228,
      "num_tokens": 11921021.0,
      "step": 5195
    },
    {
      "entropy": 5.549371528625488,
      "epoch": 0.49951969260326606,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981604217877135,
      "loss": 5.3559,
      "mean_token_accuracy": 0.19443730115890503,
      "num_tokens": 11932565.0,
      "step": 5200
    },
    {
      "entropy": 5.3739917278289795,
      "epoch": 0.5,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000498156044185496,
      "loss": 5.2818,
      "mean_token_accuracy": 0.19840114265680314,
      "num_tokens": 11943225.0,
      "step": 5205
    },
    {
      "entropy": 5.454575777053833,
      "epoch": 0.5004803073967339,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981516614022579,
      "loss": 5.2702,
      "mean_token_accuracy": 0.2014799103140831,
      "num_tokens": 11954821.0,
      "step": 5210
    },
    {
      "entropy": 5.45289101600647,
      "epoch": 0.5009606147934679,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004981472734381008,
      "loss": 5.2109,
      "mean_token_accuracy": 0.2056070551276207,
      "num_tokens": 11966090.0,
      "step": 5215
    },
    {
      "entropy": 5.432686042785645,
      "epoch": 0.5014409221902018,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981428802931267,
      "loss": 5.2493,
      "mean_token_accuracy": 0.19887453466653823,
      "num_tokens": 11977410.0,
      "step": 5220
    },
    {
      "entropy": 5.431569290161133,
      "epoch": 0.5019212295869356,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004981384819674375,
      "loss": 5.2398,
      "mean_token_accuracy": 0.19690183252096177,
      "num_tokens": 11989119.0,
      "step": 5225
    },
    {
      "entropy": 5.544884872436524,
      "epoch": 0.5024015369836695,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004981340784611354,
      "loss": 5.3326,
      "mean_token_accuracy": 0.1929663226008415,
      "num_tokens": 12000165.0,
      "step": 5230
    },
    {
      "entropy": 5.433797931671142,
      "epoch": 0.5028818443804035,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004981296697743224,
      "loss": 5.2958,
      "mean_token_accuracy": 0.1939511552453041,
      "num_tokens": 12012118.0,
      "step": 5235
    },
    {
      "entropy": 5.483579301834107,
      "epoch": 0.5033621517771374,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004981252559071012,
      "loss": 5.3412,
      "mean_token_accuracy": 0.19039848893880845,
      "num_tokens": 12023432.0,
      "step": 5240
    },
    {
      "entropy": 5.47205867767334,
      "epoch": 0.5038424591738713,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981208368595739,
      "loss": 5.24,
      "mean_token_accuracy": 0.20507996380329133,
      "num_tokens": 12034323.0,
      "step": 5245
    },
    {
      "entropy": 5.375516033172607,
      "epoch": 0.5043227665706052,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004981164126318435,
      "loss": 5.2441,
      "mean_token_accuracy": 0.1988113269209862,
      "num_tokens": 12045532.0,
      "step": 5250
    },
    {
      "entropy": 5.480970144271851,
      "epoch": 0.5048030739673391,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004981119832240124,
      "loss": 5.2675,
      "mean_token_accuracy": 0.1947346433997154,
      "num_tokens": 12057346.0,
      "step": 5255
    },
    {
      "entropy": 5.3928892612457275,
      "epoch": 0.505283381364073,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004981075486361837,
      "loss": 5.2072,
      "mean_token_accuracy": 0.2055114433169365,
      "num_tokens": 12068670.0,
      "step": 5260
    },
    {
      "entropy": 5.462854099273682,
      "epoch": 0.5057636887608069,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004981031088684601,
      "loss": 5.3435,
      "mean_token_accuracy": 0.1926040083169937,
      "num_tokens": 12079664.0,
      "step": 5265
    },
    {
      "entropy": 5.479207324981689,
      "epoch": 0.5062439961575408,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980986639209448,
      "loss": 5.2852,
      "mean_token_accuracy": 0.20990796238183976,
      "num_tokens": 12089984.0,
      "step": 5270
    },
    {
      "entropy": 5.366336822509766,
      "epoch": 0.5067243035542748,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498094213793741,
      "loss": 5.2088,
      "mean_token_accuracy": 0.20466214418411255,
      "num_tokens": 12101182.0,
      "step": 5275
    },
    {
      "entropy": 5.498898315429687,
      "epoch": 0.5072046109510087,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498089758486952,
      "loss": 5.28,
      "mean_token_accuracy": 0.1944056808948517,
      "num_tokens": 12112002.0,
      "step": 5280
    },
    {
      "entropy": 5.509644031524658,
      "epoch": 0.5076849183477425,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004980852980006812,
      "loss": 5.3716,
      "mean_token_accuracy": 0.18642902821302415,
      "num_tokens": 12124194.0,
      "step": 5285
    },
    {
      "entropy": 5.428500461578369,
      "epoch": 0.5081652257444764,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004980808323350323,
      "loss": 5.287,
      "mean_token_accuracy": 0.20239519625902175,
      "num_tokens": 12133966.0,
      "step": 5290
    },
    {
      "entropy": 5.440308284759522,
      "epoch": 0.5086455331412104,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980763614901089,
      "loss": 5.2319,
      "mean_token_accuracy": 0.20603344440460206,
      "num_tokens": 12145643.0,
      "step": 5295
    },
    {
      "entropy": 5.412059211730957,
      "epoch": 0.5091258405379443,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004980718854660146,
      "loss": 5.2843,
      "mean_token_accuracy": 0.1935652792453766,
      "num_tokens": 12156804.0,
      "step": 5300
    },
    {
      "entropy": 5.4437727451324465,
      "epoch": 0.5096061479346782,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004980674042628537,
      "loss": 5.2475,
      "mean_token_accuracy": 0.19508322924375535,
      "num_tokens": 12168700.0,
      "step": 5305
    },
    {
      "entropy": 5.47620415687561,
      "epoch": 0.5100864553314121,
      "grad_norm": 0.859375,
      "learning_rate": 0.00049806291788073,
      "loss": 5.2565,
      "mean_token_accuracy": 0.19422244131565095,
      "num_tokens": 12181050.0,
      "step": 5310
    },
    {
      "entropy": 5.502141284942627,
      "epoch": 0.510566762728146,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004980584263197477,
      "loss": 5.3387,
      "mean_token_accuracy": 0.19137953519821166,
      "num_tokens": 12192001.0,
      "step": 5315
    },
    {
      "entropy": 5.551894998550415,
      "epoch": 0.5110470701248799,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004980539295800111,
      "loss": 5.4479,
      "mean_token_accuracy": 0.1882469743490219,
      "num_tokens": 12202436.0,
      "step": 5320
    },
    {
      "entropy": 5.434849834442138,
      "epoch": 0.5115273775216138,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004980494276616246,
      "loss": 5.2412,
      "mean_token_accuracy": 0.2006586894392967,
      "num_tokens": 12214454.0,
      "step": 5325
    },
    {
      "entropy": 5.4280845642089846,
      "epoch": 0.5120076849183477,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004980449205646926,
      "loss": 5.2545,
      "mean_token_accuracy": 0.20265544056892396,
      "num_tokens": 12225924.0,
      "step": 5330
    },
    {
      "entropy": 5.50420036315918,
      "epoch": 0.5124879923150817,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00049804040828932,
      "loss": 5.2618,
      "mean_token_accuracy": 0.2032623529434204,
      "num_tokens": 12236456.0,
      "step": 5335
    },
    {
      "entropy": 5.463288831710815,
      "epoch": 0.5129682997118156,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980358908356113,
      "loss": 5.2693,
      "mean_token_accuracy": 0.20118008106946944,
      "num_tokens": 12247719.0,
      "step": 5340
    },
    {
      "entropy": 5.4230491638183596,
      "epoch": 0.5134486071085494,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004980313682036717,
      "loss": 5.3235,
      "mean_token_accuracy": 0.19510107785463332,
      "num_tokens": 12259141.0,
      "step": 5345
    },
    {
      "entropy": 5.561901187896728,
      "epoch": 0.5139289145052833,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004980268403936058,
      "loss": 5.3712,
      "mean_token_accuracy": 0.19242959171533586,
      "num_tokens": 12269748.0,
      "step": 5350
    },
    {
      "entropy": 5.478623867034912,
      "epoch": 0.5144092219020173,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980223074055189,
      "loss": 5.3164,
      "mean_token_accuracy": 0.20066264569759368,
      "num_tokens": 12281456.0,
      "step": 5355
    },
    {
      "entropy": 5.458214664459229,
      "epoch": 0.5148895292987512,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004980177692395164,
      "loss": 5.295,
      "mean_token_accuracy": 0.19240615367889405,
      "num_tokens": 12293763.0,
      "step": 5360
    },
    {
      "entropy": 5.428730916976929,
      "epoch": 0.5153698366954851,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004980132258957035,
      "loss": 5.2159,
      "mean_token_accuracy": 0.20714261382818222,
      "num_tokens": 12305398.0,
      "step": 5365
    },
    {
      "entropy": 5.371615886688232,
      "epoch": 0.515850144092219,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004980086773741856,
      "loss": 5.204,
      "mean_token_accuracy": 0.19764825254678725,
      "num_tokens": 12316582.0,
      "step": 5370
    },
    {
      "entropy": 5.510064649581909,
      "epoch": 0.516330451488953,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004980041236750685,
      "loss": 5.3154,
      "mean_token_accuracy": 0.19602434784173967,
      "num_tokens": 12328463.0,
      "step": 5375
    },
    {
      "entropy": 5.474542284011841,
      "epoch": 0.5168107588856868,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979995647984577,
      "loss": 5.3055,
      "mean_token_accuracy": 0.19989489912986755,
      "num_tokens": 12341040.0,
      "step": 5380
    },
    {
      "entropy": 5.3581665515899655,
      "epoch": 0.5172910662824207,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004979950007444593,
      "loss": 5.1985,
      "mean_token_accuracy": 0.2028449460864067,
      "num_tokens": 12353024.0,
      "step": 5385
    },
    {
      "entropy": 5.40923433303833,
      "epoch": 0.5177713736791547,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979904315131792,
      "loss": 5.2095,
      "mean_token_accuracy": 0.2023212730884552,
      "num_tokens": 12366100.0,
      "step": 5390
    },
    {
      "entropy": 5.3701718807220455,
      "epoch": 0.5182516810758886,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979858571047233,
      "loss": 5.2133,
      "mean_token_accuracy": 0.20770515501499176,
      "num_tokens": 12377829.0,
      "step": 5395
    },
    {
      "entropy": 5.512105274200439,
      "epoch": 0.5187319884726225,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979812775191979,
      "loss": 5.3204,
      "mean_token_accuracy": 0.19796691536903382,
      "num_tokens": 12390830.0,
      "step": 5400
    },
    {
      "entropy": 5.374357843399048,
      "epoch": 0.5192122958693564,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979766927567094,
      "loss": 5.1835,
      "mean_token_accuracy": 0.20305409133434296,
      "num_tokens": 12401642.0,
      "step": 5405
    },
    {
      "entropy": 5.429293298721314,
      "epoch": 0.5196926032660903,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979721028173643,
      "loss": 5.2853,
      "mean_token_accuracy": 0.1978478580713272,
      "num_tokens": 12411653.0,
      "step": 5410
    },
    {
      "entropy": 5.46315484046936,
      "epoch": 0.5201729106628242,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000497967507701269,
      "loss": 5.1935,
      "mean_token_accuracy": 0.2083654746413231,
      "num_tokens": 12422891.0,
      "step": 5415
    },
    {
      "entropy": 5.46509952545166,
      "epoch": 0.5206532180595581,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979629074085303,
      "loss": 5.2766,
      "mean_token_accuracy": 0.20137949436903,
      "num_tokens": 12434190.0,
      "step": 5420
    },
    {
      "entropy": 5.4995557308197025,
      "epoch": 0.521133525456292,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004979583019392548,
      "loss": 5.3395,
      "mean_token_accuracy": 0.19757068753242493,
      "num_tokens": 12445796.0,
      "step": 5425
    },
    {
      "entropy": 5.561598253250122,
      "epoch": 0.521613832853026,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004979536912935497,
      "loss": 5.4208,
      "mean_token_accuracy": 0.1920577585697174,
      "num_tokens": 12456212.0,
      "step": 5430
    },
    {
      "entropy": 5.361889886856079,
      "epoch": 0.5220941402497599,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497949075471522,
      "loss": 5.1194,
      "mean_token_accuracy": 0.2050475671887398,
      "num_tokens": 12467871.0,
      "step": 5435
    },
    {
      "entropy": 5.452909755706787,
      "epoch": 0.5225744476464937,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004979444544732786,
      "loss": 5.2174,
      "mean_token_accuracy": 0.1954452022910118,
      "num_tokens": 12478626.0,
      "step": 5440
    },
    {
      "entropy": 5.441848993301392,
      "epoch": 0.5230547550432276,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000497939828298927,
      "loss": 5.303,
      "mean_token_accuracy": 0.19818771183490752,
      "num_tokens": 12491487.0,
      "step": 5445
    },
    {
      "entropy": 5.503297567367554,
      "epoch": 0.5235350624399616,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979351969485747,
      "loss": 5.2575,
      "mean_token_accuracy": 0.1951201155781746,
      "num_tokens": 12503240.0,
      "step": 5450
    },
    {
      "entropy": 5.464956569671631,
      "epoch": 0.5240153698366955,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979305604223291,
      "loss": 5.2159,
      "mean_token_accuracy": 0.19913747459650039,
      "num_tokens": 12513860.0,
      "step": 5455
    },
    {
      "entropy": 5.38545298576355,
      "epoch": 0.5244956772334294,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979259187202978,
      "loss": 5.2855,
      "mean_token_accuracy": 0.19753282517194748,
      "num_tokens": 12525884.0,
      "step": 5460
    },
    {
      "entropy": 5.465018606185913,
      "epoch": 0.5249759846301633,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004979212718425887,
      "loss": 5.2025,
      "mean_token_accuracy": 0.19672263711690902,
      "num_tokens": 12536709.0,
      "step": 5465
    },
    {
      "entropy": 5.452314710617065,
      "epoch": 0.5254562920268973,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979166197893096,
      "loss": 5.1747,
      "mean_token_accuracy": 0.2059873014688492,
      "num_tokens": 12549727.0,
      "step": 5470
    },
    {
      "entropy": 5.409617853164673,
      "epoch": 0.5259365994236311,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004979119625605683,
      "loss": 5.2703,
      "mean_token_accuracy": 0.1943661794066429,
      "num_tokens": 12562053.0,
      "step": 5475
    },
    {
      "entropy": 5.334145212173462,
      "epoch": 0.526416906820365,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979073001564734,
      "loss": 5.1597,
      "mean_token_accuracy": 0.2090065971016884,
      "num_tokens": 12574096.0,
      "step": 5480
    },
    {
      "entropy": 5.446299457550049,
      "epoch": 0.5268972142170989,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979026325771328,
      "loss": 5.3383,
      "mean_token_accuracy": 0.19677683115005493,
      "num_tokens": 12585416.0,
      "step": 5485
    },
    {
      "entropy": 5.424954462051391,
      "epoch": 0.5273775216138329,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004978979598226549,
      "loss": 5.1861,
      "mean_token_accuracy": 0.2083905816078186,
      "num_tokens": 12596861.0,
      "step": 5490
    },
    {
      "entropy": 5.364092206954956,
      "epoch": 0.5278578290105668,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978932818931483,
      "loss": 5.2039,
      "mean_token_accuracy": 0.20646013617515563,
      "num_tokens": 12607761.0,
      "step": 5495
    },
    {
      "entropy": 5.488156509399414,
      "epoch": 0.5283381364073007,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978885987887216,
      "loss": 5.3187,
      "mean_token_accuracy": 0.19835564345121384,
      "num_tokens": 12619889.0,
      "step": 5500
    },
    {
      "entropy": 5.516491365432739,
      "epoch": 0.5288184438040345,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978839105094833,
      "loss": 5.3012,
      "mean_token_accuracy": 0.19845593124628066,
      "num_tokens": 12630604.0,
      "step": 5505
    },
    {
      "entropy": 5.354294157028198,
      "epoch": 0.5292987512007685,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978792170555426,
      "loss": 5.1826,
      "mean_token_accuracy": 0.20391630083322526,
      "num_tokens": 12641172.0,
      "step": 5510
    },
    {
      "entropy": 5.373248291015625,
      "epoch": 0.5297790585975024,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978745184270083,
      "loss": 5.1094,
      "mean_token_accuracy": 0.2145618200302124,
      "num_tokens": 12651731.0,
      "step": 5515
    },
    {
      "entropy": 5.428430986404419,
      "epoch": 0.5302593659942363,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978698146239893,
      "loss": 5.2315,
      "mean_token_accuracy": 0.19880771040916442,
      "num_tokens": 12663050.0,
      "step": 5520
    },
    {
      "entropy": 5.507349920272827,
      "epoch": 0.5307396733909702,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978651056465952,
      "loss": 5.336,
      "mean_token_accuracy": 0.19323476850986482,
      "num_tokens": 12674732.0,
      "step": 5525
    },
    {
      "entropy": 5.358083295822143,
      "epoch": 0.5312199807877042,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497860391494935,
      "loss": 5.1604,
      "mean_token_accuracy": 0.20278584212064743,
      "num_tokens": 12685981.0,
      "step": 5530
    },
    {
      "entropy": 5.474340152740479,
      "epoch": 0.531700288184438,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978556721691183,
      "loss": 5.291,
      "mean_token_accuracy": 0.19821719378232955,
      "num_tokens": 12697139.0,
      "step": 5535
    },
    {
      "entropy": 5.382588005065918,
      "epoch": 0.5321805955811719,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978509476692547,
      "loss": 5.2361,
      "mean_token_accuracy": 0.19628128707408904,
      "num_tokens": 12708268.0,
      "step": 5540
    },
    {
      "entropy": 5.412813377380371,
      "epoch": 0.5326609029779059,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004978462179954538,
      "loss": 5.2185,
      "mean_token_accuracy": 0.20490823090076446,
      "num_tokens": 12720715.0,
      "step": 5545
    },
    {
      "entropy": 5.413670349121094,
      "epoch": 0.5331412103746398,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978414831478253,
      "loss": 5.1951,
      "mean_token_accuracy": 0.20597660094499587,
      "num_tokens": 12732409.0,
      "step": 5550
    },
    {
      "entropy": 5.36975474357605,
      "epoch": 0.5336215177713737,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978367431264794,
      "loss": 5.3292,
      "mean_token_accuracy": 0.19064888060092927,
      "num_tokens": 12745174.0,
      "step": 5555
    },
    {
      "entropy": 5.495306825637817,
      "epoch": 0.5341018251681076,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004978319979315261,
      "loss": 5.2716,
      "mean_token_accuracy": 0.19742067754268647,
      "num_tokens": 12756116.0,
      "step": 5560
    },
    {
      "entropy": 5.466085290908813,
      "epoch": 0.5345821325648416,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978272475630752,
      "loss": 5.2048,
      "mean_token_accuracy": 0.20205890983343125,
      "num_tokens": 12768183.0,
      "step": 5565
    },
    {
      "entropy": 5.321042108535766,
      "epoch": 0.5350624399615754,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004978224920212374,
      "loss": 5.226,
      "mean_token_accuracy": 0.1981593042612076,
      "num_tokens": 12778537.0,
      "step": 5570
    },
    {
      "entropy": 5.383537244796753,
      "epoch": 0.5355427473583093,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978177313061232,
      "loss": 5.2597,
      "mean_token_accuracy": 0.19882985055446625,
      "num_tokens": 12789691.0,
      "step": 5575
    },
    {
      "entropy": 5.494786214828491,
      "epoch": 0.5360230547550432,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978129654178426,
      "loss": 5.281,
      "mean_token_accuracy": 0.19676683098077774,
      "num_tokens": 12801438.0,
      "step": 5580
    },
    {
      "entropy": 5.452530431747436,
      "epoch": 0.5365033621517772,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978081943565067,
      "loss": 5.2303,
      "mean_token_accuracy": 0.19537653774023056,
      "num_tokens": 12812425.0,
      "step": 5585
    },
    {
      "entropy": 5.37575855255127,
      "epoch": 0.5369836695485111,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004978034181222261,
      "loss": 5.2057,
      "mean_token_accuracy": 0.19847560077905654,
      "num_tokens": 12824735.0,
      "step": 5590
    },
    {
      "entropy": 5.488990831375122,
      "epoch": 0.537463976945245,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977986367151119,
      "loss": 5.2972,
      "mean_token_accuracy": 0.20113806277513505,
      "num_tokens": 12835454.0,
      "step": 5595
    },
    {
      "entropy": 5.53157114982605,
      "epoch": 0.5379442843419788,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004977938501352747,
      "loss": 5.346,
      "mean_token_accuracy": 0.19288862645626068,
      "num_tokens": 12847086.0,
      "step": 5600
    },
    {
      "entropy": 5.461941051483154,
      "epoch": 0.5384245917387128,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004977890583828259,
      "loss": 5.2786,
      "mean_token_accuracy": 0.20097937434911728,
      "num_tokens": 12857713.0,
      "step": 5605
    },
    {
      "entropy": 5.379673480987549,
      "epoch": 0.5389048991354467,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004977842614578768,
      "loss": 5.2624,
      "mean_token_accuracy": 0.2018478959798813,
      "num_tokens": 12869967.0,
      "step": 5610
    },
    {
      "entropy": 5.492646980285644,
      "epoch": 0.5393852065321806,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977794593605386,
      "loss": 5.2923,
      "mean_token_accuracy": 0.19776214063167571,
      "num_tokens": 12881230.0,
      "step": 5615
    },
    {
      "entropy": 5.448351192474365,
      "epoch": 0.5398655139289145,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497774652090923,
      "loss": 5.2605,
      "mean_token_accuracy": 0.2008052185177803,
      "num_tokens": 12892376.0,
      "step": 5620
    },
    {
      "entropy": 5.450916433334351,
      "epoch": 0.5403458213256485,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977698396491414,
      "loss": 5.2683,
      "mean_token_accuracy": 0.2019094258546829,
      "num_tokens": 12903709.0,
      "step": 5625
    },
    {
      "entropy": 5.377445697784424,
      "epoch": 0.5408261287223823,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977650220353055,
      "loss": 5.0969,
      "mean_token_accuracy": 0.20498238056898116,
      "num_tokens": 12914958.0,
      "step": 5630
    },
    {
      "entropy": 5.372557878494263,
      "epoch": 0.5413064361191162,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977601992495274,
      "loss": 5.2213,
      "mean_token_accuracy": 0.20101455003023147,
      "num_tokens": 12927418.0,
      "step": 5635
    },
    {
      "entropy": 5.437803649902344,
      "epoch": 0.5417867435158501,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004977553712919189,
      "loss": 5.2651,
      "mean_token_accuracy": 0.2012587755918503,
      "num_tokens": 12939874.0,
      "step": 5640
    },
    {
      "entropy": 5.488529539108276,
      "epoch": 0.5422670509125841,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977505381625921,
      "loss": 5.2811,
      "mean_token_accuracy": 0.19846536368131637,
      "num_tokens": 12951113.0,
      "step": 5645
    },
    {
      "entropy": 5.358415174484253,
      "epoch": 0.542747358309318,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977456998616593,
      "loss": 5.178,
      "mean_token_accuracy": 0.20335119515657424,
      "num_tokens": 12961940.0,
      "step": 5650
    },
    {
      "entropy": 5.318795442581177,
      "epoch": 0.5432276657060519,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977408563892327,
      "loss": 5.1825,
      "mean_token_accuracy": 0.19999839067459108,
      "num_tokens": 12973938.0,
      "step": 5655
    },
    {
      "entropy": 5.446343898773193,
      "epoch": 0.5437079731027857,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004977360077454249,
      "loss": 5.1987,
      "mean_token_accuracy": 0.20178124606609343,
      "num_tokens": 12985400.0,
      "step": 5660
    },
    {
      "entropy": 5.417103862762451,
      "epoch": 0.5441882804995197,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977311539303483,
      "loss": 5.2407,
      "mean_token_accuracy": 0.20964264422655104,
      "num_tokens": 12996402.0,
      "step": 5665
    },
    {
      "entropy": 5.344844722747803,
      "epoch": 0.5446685878962536,
      "grad_norm": 1.0,
      "learning_rate": 0.0004977262949441158,
      "loss": 5.1284,
      "mean_token_accuracy": 0.21215670555830002,
      "num_tokens": 13006991.0,
      "step": 5670
    },
    {
      "entropy": 5.425043201446533,
      "epoch": 0.5451488952929875,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004977214307868399,
      "loss": 5.2295,
      "mean_token_accuracy": 0.20265681445598602,
      "num_tokens": 13016969.0,
      "step": 5675
    },
    {
      "entropy": 5.384352016448974,
      "epoch": 0.5456292026897214,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497716561458634,
      "loss": 5.1603,
      "mean_token_accuracy": 0.20864049792289735,
      "num_tokens": 13027759.0,
      "step": 5680
    },
    {
      "entropy": 5.512132787704468,
      "epoch": 0.5461095100864554,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977116869596107,
      "loss": 5.3854,
      "mean_token_accuracy": 0.1933292031288147,
      "num_tokens": 13039881.0,
      "step": 5685
    },
    {
      "entropy": 5.450646924972534,
      "epoch": 0.5465898174831892,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004977068072898834,
      "loss": 5.2281,
      "mean_token_accuracy": 0.2028298258781433,
      "num_tokens": 13051443.0,
      "step": 5690
    },
    {
      "entropy": 5.4808813571929935,
      "epoch": 0.5470701248799231,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004977019224495652,
      "loss": 5.3124,
      "mean_token_accuracy": 0.19113280028104782,
      "num_tokens": 13063474.0,
      "step": 5695
    },
    {
      "entropy": 5.38072190284729,
      "epoch": 0.547550432276657,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976970324387698,
      "loss": 5.1591,
      "mean_token_accuracy": 0.21121865510940552,
      "num_tokens": 13074365.0,
      "step": 5700
    },
    {
      "entropy": 5.384839630126953,
      "epoch": 0.548030739673391,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976921372576104,
      "loss": 5.2279,
      "mean_token_accuracy": 0.20203657448291779,
      "num_tokens": 13087354.0,
      "step": 5705
    },
    {
      "entropy": 5.395923471450805,
      "epoch": 0.5485110470701249,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004976872369062011,
      "loss": 5.2118,
      "mean_token_accuracy": 0.2014988273382187,
      "num_tokens": 13099306.0,
      "step": 5710
    },
    {
      "entropy": 5.477517414093017,
      "epoch": 0.5489913544668588,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004976823313846552,
      "loss": 5.3482,
      "mean_token_accuracy": 0.19400807470083237,
      "num_tokens": 13111259.0,
      "step": 5715
    },
    {
      "entropy": 5.509419107437134,
      "epoch": 0.5494716618635928,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976774206930869,
      "loss": 5.261,
      "mean_token_accuracy": 0.19410279840230943,
      "num_tokens": 13123589.0,
      "step": 5720
    },
    {
      "entropy": 5.385268211364746,
      "epoch": 0.5499519692603266,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004976725048316101,
      "loss": 5.2514,
      "mean_token_accuracy": 0.19971895813941956,
      "num_tokens": 13136485.0,
      "step": 5725
    },
    {
      "entropy": 5.445385360717774,
      "epoch": 0.5504322766570605,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004976675838003388,
      "loss": 5.2307,
      "mean_token_accuracy": 0.1987402245402336,
      "num_tokens": 13148067.0,
      "step": 5730
    },
    {
      "entropy": 5.429761600494385,
      "epoch": 0.5509125840537944,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004976626575993877,
      "loss": 5.2259,
      "mean_token_accuracy": 0.19749991297721864,
      "num_tokens": 13159813.0,
      "step": 5735
    },
    {
      "entropy": 5.470585727691651,
      "epoch": 0.5513928914505284,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004976577262288705,
      "loss": 5.2851,
      "mean_token_accuracy": 0.20123465061187745,
      "num_tokens": 13170828.0,
      "step": 5740
    },
    {
      "entropy": 5.350544357299805,
      "epoch": 0.5518731988472623,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004976527896889023,
      "loss": 5.1144,
      "mean_token_accuracy": 0.21216228753328323,
      "num_tokens": 13181883.0,
      "step": 5745
    },
    {
      "entropy": 5.395107269287109,
      "epoch": 0.5523535062439962,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976478479795974,
      "loss": 5.1864,
      "mean_token_accuracy": 0.20682404637336732,
      "num_tokens": 13193530.0,
      "step": 5750
    },
    {
      "entropy": 5.518691444396973,
      "epoch": 0.55283381364073,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004976429011010706,
      "loss": 5.4119,
      "mean_token_accuracy": 0.18553577214479447,
      "num_tokens": 13205822.0,
      "step": 5755
    },
    {
      "entropy": 5.390845012664795,
      "epoch": 0.553314121037464,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004976379490534366,
      "loss": 5.1375,
      "mean_token_accuracy": 0.2063609778881073,
      "num_tokens": 13216698.0,
      "step": 5760
    },
    {
      "entropy": 5.372720384597779,
      "epoch": 0.5537944284341979,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004976329918368107,
      "loss": 5.2362,
      "mean_token_accuracy": 0.20510985404253007,
      "num_tokens": 13228389.0,
      "step": 5765
    },
    {
      "entropy": 5.459476041793823,
      "epoch": 0.5542747358309318,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976280294513079,
      "loss": 5.2863,
      "mean_token_accuracy": 0.19380225241184235,
      "num_tokens": 13239628.0,
      "step": 5770
    },
    {
      "entropy": 5.45081353187561,
      "epoch": 0.5547550432276657,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004976230618970431,
      "loss": 5.2925,
      "mean_token_accuracy": 0.20101425051689148,
      "num_tokens": 13251149.0,
      "step": 5775
    },
    {
      "entropy": 5.557024431228638,
      "epoch": 0.5552353506243997,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000497618089174132,
      "loss": 5.3309,
      "mean_token_accuracy": 0.19553734362125397,
      "num_tokens": 13264846.0,
      "step": 5780
    },
    {
      "entropy": 5.340891885757446,
      "epoch": 0.5557156580211335,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004976131112826898,
      "loss": 5.1102,
      "mean_token_accuracy": 0.2120337665081024,
      "num_tokens": 13275409.0,
      "step": 5785
    },
    {
      "entropy": 5.256704425811767,
      "epoch": 0.5561959654178674,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004976081282228323,
      "loss": 5.0997,
      "mean_token_accuracy": 0.21279428601264955,
      "num_tokens": 13287173.0,
      "step": 5790
    },
    {
      "entropy": 5.466112899780273,
      "epoch": 0.5566762728146013,
      "grad_norm": 0.90625,
      "learning_rate": 0.000497603139994675,
      "loss": 5.1768,
      "mean_token_accuracy": 0.20267048478126526,
      "num_tokens": 13298225.0,
      "step": 5795
    },
    {
      "entropy": 5.411417007446289,
      "epoch": 0.5571565802113353,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975981465983338,
      "loss": 5.1918,
      "mean_token_accuracy": 0.2040710151195526,
      "num_tokens": 13309685.0,
      "step": 5800
    },
    {
      "entropy": 5.349876117706299,
      "epoch": 0.5576368876080692,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975931480339246,
      "loss": 5.201,
      "mean_token_accuracy": 0.20868574678897858,
      "num_tokens": 13320837.0,
      "step": 5805
    },
    {
      "entropy": 5.454625225067138,
      "epoch": 0.5581171950048031,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004975881443015635,
      "loss": 5.2912,
      "mean_token_accuracy": 0.1992785006761551,
      "num_tokens": 13333512.0,
      "step": 5810
    },
    {
      "entropy": 5.4970136165618895,
      "epoch": 0.5585975024015369,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004975831354013667,
      "loss": 5.3187,
      "mean_token_accuracy": 0.19946846067905427,
      "num_tokens": 13345189.0,
      "step": 5815
    },
    {
      "entropy": 5.389864587783814,
      "epoch": 0.5590778097982709,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004975781213334503,
      "loss": 5.1908,
      "mean_token_accuracy": 0.20144879668951035,
      "num_tokens": 13356123.0,
      "step": 5820
    },
    {
      "entropy": 5.391473960876465,
      "epoch": 0.5595581171950048,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004975731020979309,
      "loss": 5.2333,
      "mean_token_accuracy": 0.2022359102964401,
      "num_tokens": 13366902.0,
      "step": 5825
    },
    {
      "entropy": 5.475438785552979,
      "epoch": 0.5600384245917387,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004975680776949249,
      "loss": 5.273,
      "mean_token_accuracy": 0.19751251935958863,
      "num_tokens": 13377567.0,
      "step": 5830
    },
    {
      "entropy": 5.451048517227173,
      "epoch": 0.5605187319884726,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975630481245492,
      "loss": 5.2281,
      "mean_token_accuracy": 0.20986885875463485,
      "num_tokens": 13387297.0,
      "step": 5835
    },
    {
      "entropy": 5.444889259338379,
      "epoch": 0.5609990393852066,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975580133869202,
      "loss": 5.2529,
      "mean_token_accuracy": 0.20367213785648347,
      "num_tokens": 13397723.0,
      "step": 5840
    },
    {
      "entropy": 5.440711307525635,
      "epoch": 0.5614793467819404,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975529734821552,
      "loss": 5.3125,
      "mean_token_accuracy": 0.19717175215482713,
      "num_tokens": 13409875.0,
      "step": 5845
    },
    {
      "entropy": 5.372532749176026,
      "epoch": 0.5619596541786743,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004975479284103708,
      "loss": 5.2079,
      "mean_token_accuracy": 0.20235518962144852,
      "num_tokens": 13421338.0,
      "step": 5850
    },
    {
      "entropy": 5.458660697937011,
      "epoch": 0.5624399615754082,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004975428781716845,
      "loss": 5.2462,
      "mean_token_accuracy": 0.19968458116054535,
      "num_tokens": 13431373.0,
      "step": 5855
    },
    {
      "entropy": 5.4430427074432375,
      "epoch": 0.5629202689721422,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004975378227662134,
      "loss": 5.2574,
      "mean_token_accuracy": 0.2052182987332344,
      "num_tokens": 13443158.0,
      "step": 5860
    },
    {
      "entropy": 5.493396043777466,
      "epoch": 0.5634005763688761,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004975327621940746,
      "loss": 5.3251,
      "mean_token_accuracy": 0.19516831785440444,
      "num_tokens": 13454559.0,
      "step": 5865
    },
    {
      "entropy": 5.5587399959564205,
      "epoch": 0.56388088376561,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004975276964553861,
      "loss": 5.3882,
      "mean_token_accuracy": 0.19617153257131575,
      "num_tokens": 13466934.0,
      "step": 5870
    },
    {
      "entropy": 5.428758668899536,
      "epoch": 0.5643611911623438,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004975226255502651,
      "loss": 5.1403,
      "mean_token_accuracy": 0.21330775022506715,
      "num_tokens": 13477770.0,
      "step": 5875
    },
    {
      "entropy": 5.4610990524292,
      "epoch": 0.5648414985590778,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004975175494788297,
      "loss": 5.3431,
      "mean_token_accuracy": 0.19222095012664794,
      "num_tokens": 13490093.0,
      "step": 5880
    },
    {
      "entropy": 5.52604112625122,
      "epoch": 0.5653218059558117,
      "grad_norm": 1.25,
      "learning_rate": 0.0004975124682411974,
      "loss": 5.2249,
      "mean_token_accuracy": 0.20205526053905487,
      "num_tokens": 13500663.0,
      "step": 5885
    },
    {
      "entropy": 5.442627429962158,
      "epoch": 0.5658021133525456,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975073818374863,
      "loss": 5.3041,
      "mean_token_accuracy": 0.1966984361410141,
      "num_tokens": 13512369.0,
      "step": 5890
    },
    {
      "entropy": 5.422051334381104,
      "epoch": 0.5662824207492796,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004975022902678145,
      "loss": 5.1741,
      "mean_token_accuracy": 0.1990967184305191,
      "num_tokens": 13523181.0,
      "step": 5895
    },
    {
      "entropy": 5.411604356765747,
      "epoch": 0.5667627281460135,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004974971935323003,
      "loss": 5.1446,
      "mean_token_accuracy": 0.19783616960048675,
      "num_tokens": 13534113.0,
      "step": 5900
    },
    {
      "entropy": 5.355008316040039,
      "epoch": 0.5672430355427474,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004974920916310619,
      "loss": 5.188,
      "mean_token_accuracy": 0.20168602615594863,
      "num_tokens": 13545037.0,
      "step": 5905
    },
    {
      "entropy": 5.4220489978790285,
      "epoch": 0.5677233429394812,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974869845642178,
      "loss": 5.2284,
      "mean_token_accuracy": 0.19897332787513733,
      "num_tokens": 13555541.0,
      "step": 5910
    },
    {
      "entropy": 5.389899873733521,
      "epoch": 0.5682036503362152,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004974818723318866,
      "loss": 5.2105,
      "mean_token_accuracy": 0.20374743044376373,
      "num_tokens": 13566951.0,
      "step": 5915
    },
    {
      "entropy": 5.459915590286255,
      "epoch": 0.5686839577329491,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974767549341868,
      "loss": 5.287,
      "mean_token_accuracy": 0.19948874711990355,
      "num_tokens": 13578492.0,
      "step": 5920
    },
    {
      "entropy": 5.4552897930145265,
      "epoch": 0.569164265129683,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004974716323712376,
      "loss": 5.183,
      "mean_token_accuracy": 0.2121192216873169,
      "num_tokens": 13589183.0,
      "step": 5925
    },
    {
      "entropy": 5.458286571502685,
      "epoch": 0.5696445725264169,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974665046431576,
      "loss": 5.3108,
      "mean_token_accuracy": 0.19659883230924607,
      "num_tokens": 13600588.0,
      "step": 5930
    },
    {
      "entropy": 5.3862463474273685,
      "epoch": 0.5701248799231509,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974613717500659,
      "loss": 5.1979,
      "mean_token_accuracy": 0.20755694359540938,
      "num_tokens": 13612107.0,
      "step": 5935
    },
    {
      "entropy": 5.540802001953125,
      "epoch": 0.5706051873198847,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974562336920818,
      "loss": 5.3507,
      "mean_token_accuracy": 0.19629850685596467,
      "num_tokens": 13623973.0,
      "step": 5940
    },
    {
      "entropy": 5.400304937362671,
      "epoch": 0.5710854947166186,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004974510904693245,
      "loss": 5.161,
      "mean_token_accuracy": 0.2056756779551506,
      "num_tokens": 13634994.0,
      "step": 5945
    },
    {
      "entropy": 5.467379760742188,
      "epoch": 0.5715658021133525,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004974459420819134,
      "loss": 5.3338,
      "mean_token_accuracy": 0.1981966733932495,
      "num_tokens": 13646361.0,
      "step": 5950
    },
    {
      "entropy": 5.423090648651123,
      "epoch": 0.5720461095100865,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497440788529968,
      "loss": 5.2022,
      "mean_token_accuracy": 0.196497642993927,
      "num_tokens": 13656975.0,
      "step": 5955
    },
    {
      "entropy": 5.415749454498291,
      "epoch": 0.5725264169068204,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974356298136081,
      "loss": 5.2613,
      "mean_token_accuracy": 0.1987801671028137,
      "num_tokens": 13668434.0,
      "step": 5960
    },
    {
      "entropy": 5.467151117324829,
      "epoch": 0.5730067243035543,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004974304659329533,
      "loss": 5.2347,
      "mean_token_accuracy": 0.20059672445058824,
      "num_tokens": 13679266.0,
      "step": 5965
    },
    {
      "entropy": 5.36699423789978,
      "epoch": 0.5734870317002881,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004974252968881236,
      "loss": 5.246,
      "mean_token_accuracy": 0.19388700574636458,
      "num_tokens": 13690921.0,
      "step": 5970
    },
    {
      "entropy": 5.427312994003296,
      "epoch": 0.5739673390970221,
      "grad_norm": 0.890625,
      "learning_rate": 0.000497420122679239,
      "loss": 5.2075,
      "mean_token_accuracy": 0.20325016975402832,
      "num_tokens": 13702329.0,
      "step": 5975
    },
    {
      "entropy": 5.384410715103149,
      "epoch": 0.574447646493756,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004974149433064196,
      "loss": 5.1722,
      "mean_token_accuracy": 0.21002744883298874,
      "num_tokens": 13713356.0,
      "step": 5980
    },
    {
      "entropy": 5.311866712570191,
      "epoch": 0.5749279538904899,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004974097587697856,
      "loss": 5.1492,
      "mean_token_accuracy": 0.20121487677097322,
      "num_tokens": 13724718.0,
      "step": 5985
    },
    {
      "entropy": 5.310837936401367,
      "epoch": 0.5754082612872238,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004974045690694575,
      "loss": 5.1827,
      "mean_token_accuracy": 0.20771329551935197,
      "num_tokens": 13736113.0,
      "step": 5990
    },
    {
      "entropy": 5.484830284118653,
      "epoch": 0.5758885686839578,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004973993742055557,
      "loss": 5.2038,
      "mean_token_accuracy": 0.20306566208600998,
      "num_tokens": 13748322.0,
      "step": 5995
    },
    {
      "entropy": 5.364419174194336,
      "epoch": 0.5763688760806917,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004973941741782007,
      "loss": 5.2097,
      "mean_token_accuracy": 0.19368440806865692,
      "num_tokens": 13759433.0,
      "step": 6000
    },
    {
      "epoch": 0.5763688760806917,
      "eval_entropy": 5.168700758740124,
      "eval_loss": 5.236293315887451,
      "eval_mean_token_accuracy": 0.20920352415176702,
      "eval_num_tokens": 13759433.0,
      "eval_runtime": 20.4353,
      "eval_samples_per_second": 1605.798,
      "eval_steps_per_second": 200.731,
      "step": 6000
    },
    {
      "entropy": 5.329404592514038,
      "epoch": 0.5768491834774255,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004973889689875135,
      "loss": 5.1332,
      "mean_token_accuracy": 0.20637214481830596,
      "num_tokens": 13770181.0,
      "step": 6005
    },
    {
      "entropy": 5.371566581726074,
      "epoch": 0.5773294908741594,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973837586336147,
      "loss": 5.1748,
      "mean_token_accuracy": 0.20478294044733047,
      "num_tokens": 13781792.0,
      "step": 6010
    },
    {
      "entropy": 5.420533657073975,
      "epoch": 0.5778097982708934,
      "grad_norm": 0.875,
      "learning_rate": 0.0004973785431166254,
      "loss": 5.2313,
      "mean_token_accuracy": 0.195565801858902,
      "num_tokens": 13792101.0,
      "step": 6015
    },
    {
      "entropy": 5.406401586532593,
      "epoch": 0.5782901056676273,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973733224366666,
      "loss": 5.2301,
      "mean_token_accuracy": 0.20410930812358857,
      "num_tokens": 13803640.0,
      "step": 6020
    },
    {
      "entropy": 5.417542219161987,
      "epoch": 0.5787704130643612,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973680965938597,
      "loss": 5.1536,
      "mean_token_accuracy": 0.20339121520519257,
      "num_tokens": 13815017.0,
      "step": 6025
    },
    {
      "entropy": 5.495737361907959,
      "epoch": 0.579250720461095,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004973628655883258,
      "loss": 5.3677,
      "mean_token_accuracy": 0.1945173442363739,
      "num_tokens": 13826119.0,
      "step": 6030
    },
    {
      "entropy": 5.38583025932312,
      "epoch": 0.579731027857829,
      "grad_norm": 1.0,
      "learning_rate": 0.0004973576294201865,
      "loss": 5.1854,
      "mean_token_accuracy": 0.20032606571912764,
      "num_tokens": 13837869.0,
      "step": 6035
    },
    {
      "entropy": 5.602328157424926,
      "epoch": 0.5802113352545629,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973523880895633,
      "loss": 5.3404,
      "mean_token_accuracy": 0.19249229580163957,
      "num_tokens": 13849333.0,
      "step": 6040
    },
    {
      "entropy": 5.459065675735474,
      "epoch": 0.5806916426512968,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004973471415965779,
      "loss": 5.2107,
      "mean_token_accuracy": 0.20107725411653518,
      "num_tokens": 13860648.0,
      "step": 6045
    },
    {
      "entropy": 5.476138162612915,
      "epoch": 0.5811719500480308,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497341889941352,
      "loss": 5.3442,
      "mean_token_accuracy": 0.19328550547361373,
      "num_tokens": 13872979.0,
      "step": 6050
    },
    {
      "entropy": 5.442042350769043,
      "epoch": 0.5816522574447647,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004973366331240078,
      "loss": 5.2314,
      "mean_token_accuracy": 0.1992579609155655,
      "num_tokens": 13884363.0,
      "step": 6055
    },
    {
      "entropy": 5.404986906051636,
      "epoch": 0.5821325648414986,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000497331371144667,
      "loss": 5.2123,
      "mean_token_accuracy": 0.1971079334616661,
      "num_tokens": 13895424.0,
      "step": 6060
    },
    {
      "entropy": 5.436076259613037,
      "epoch": 0.5826128722382324,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004973261040034521,
      "loss": 5.3382,
      "mean_token_accuracy": 0.1891995221376419,
      "num_tokens": 13907319.0,
      "step": 6065
    },
    {
      "entropy": 5.334365272521973,
      "epoch": 0.5830931796349664,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004973208317004852,
      "loss": 5.1769,
      "mean_token_accuracy": 0.20466973632574081,
      "num_tokens": 13920013.0,
      "step": 6070
    },
    {
      "entropy": 5.403887748718262,
      "epoch": 0.5835734870317003,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004973155542358889,
      "loss": 5.1526,
      "mean_token_accuracy": 0.20059441030025482,
      "num_tokens": 13932033.0,
      "step": 6075
    },
    {
      "entropy": 5.368381643295288,
      "epoch": 0.5840537944284342,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973102716097853,
      "loss": 5.1559,
      "mean_token_accuracy": 0.20629478245973587,
      "num_tokens": 13943324.0,
      "step": 6080
    },
    {
      "entropy": 5.409040641784668,
      "epoch": 0.5845341018251681,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973049838222973,
      "loss": 5.235,
      "mean_token_accuracy": 0.2110427588224411,
      "num_tokens": 13954291.0,
      "step": 6085
    },
    {
      "entropy": 5.445193576812744,
      "epoch": 0.5850144092219021,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004972996908735479,
      "loss": 5.3392,
      "mean_token_accuracy": 0.19406466633081437,
      "num_tokens": 13966264.0,
      "step": 6090
    },
    {
      "entropy": 5.504749345779419,
      "epoch": 0.585494716618636,
      "grad_norm": 0.875,
      "learning_rate": 0.0004972943927636597,
      "loss": 5.2221,
      "mean_token_accuracy": 0.19688427746295928,
      "num_tokens": 13977785.0,
      "step": 6095
    },
    {
      "entropy": 5.351817083358765,
      "epoch": 0.5859750240153698,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004972890894927558,
      "loss": 5.1975,
      "mean_token_accuracy": 0.20503768771886827,
      "num_tokens": 13989704.0,
      "step": 6100
    },
    {
      "entropy": 5.399487400054932,
      "epoch": 0.5864553314121037,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004972837810609592,
      "loss": 5.1855,
      "mean_token_accuracy": 0.20075137615203859,
      "num_tokens": 14000565.0,
      "step": 6105
    },
    {
      "entropy": 5.327906084060669,
      "epoch": 0.5869356388088377,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004972784674683933,
      "loss": 5.1178,
      "mean_token_accuracy": 0.20848890542984008,
      "num_tokens": 14011521.0,
      "step": 6110
    },
    {
      "entropy": 5.3874894142150875,
      "epoch": 0.5874159462055716,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004972731487151815,
      "loss": 5.1451,
      "mean_token_accuracy": 0.21039289683103563,
      "num_tokens": 14022966.0,
      "step": 6115
    },
    {
      "entropy": 5.497272634506226,
      "epoch": 0.5878962536023055,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004972678248014471,
      "loss": 5.2751,
      "mean_token_accuracy": 0.20130295306444168,
      "num_tokens": 14034905.0,
      "step": 6120
    },
    {
      "entropy": 5.3000682353973385,
      "epoch": 0.5883765609990393,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004972624957273139,
      "loss": 5.1274,
      "mean_token_accuracy": 0.20796251744031907,
      "num_tokens": 14045816.0,
      "step": 6125
    },
    {
      "entropy": 5.326866817474365,
      "epoch": 0.5888568683957733,
      "grad_norm": 0.875,
      "learning_rate": 0.0004972571614929055,
      "loss": 5.1034,
      "mean_token_accuracy": 0.20375093966722488,
      "num_tokens": 14057316.0,
      "step": 6130
    },
    {
      "entropy": 5.451218509674073,
      "epoch": 0.5893371757925072,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004972518220983457,
      "loss": 5.2453,
      "mean_token_accuracy": 0.1933143064379692,
      "num_tokens": 14067542.0,
      "step": 6135
    },
    {
      "entropy": 5.339518451690674,
      "epoch": 0.5898174831892411,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004972464775437586,
      "loss": 5.1006,
      "mean_token_accuracy": 0.20812444388866425,
      "num_tokens": 14079467.0,
      "step": 6140
    },
    {
      "entropy": 5.342601633071899,
      "epoch": 0.590297790585975,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004972411278292683,
      "loss": 5.0578,
      "mean_token_accuracy": 0.21310915350914,
      "num_tokens": 14090695.0,
      "step": 6145
    },
    {
      "entropy": 5.366895914077759,
      "epoch": 0.590778097982709,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004972357729549988,
      "loss": 5.1175,
      "mean_token_accuracy": 0.21252727508544922,
      "num_tokens": 14101380.0,
      "step": 6150
    },
    {
      "entropy": 5.332993078231811,
      "epoch": 0.5912584053794429,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004972304129210746,
      "loss": 5.151,
      "mean_token_accuracy": 0.21010346561670304,
      "num_tokens": 14112265.0,
      "step": 6155
    },
    {
      "entropy": 5.397362756729126,
      "epoch": 0.5917387127761767,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004972250477276202,
      "loss": 5.158,
      "mean_token_accuracy": 0.2071602910757065,
      "num_tokens": 14124294.0,
      "step": 6160
    },
    {
      "entropy": 5.391251707077027,
      "epoch": 0.5922190201729106,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972196773747599,
      "loss": 5.1853,
      "mean_token_accuracy": 0.20393324494361878,
      "num_tokens": 14135230.0,
      "step": 6165
    },
    {
      "entropy": 5.432328605651856,
      "epoch": 0.5926993275696446,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004972143018626186,
      "loss": 5.273,
      "mean_token_accuracy": 0.1966063916683197,
      "num_tokens": 14147576.0,
      "step": 6170
    },
    {
      "entropy": 5.36362590789795,
      "epoch": 0.5931796349663785,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004972089211913211,
      "loss": 5.1581,
      "mean_token_accuracy": 0.20785517245531082,
      "num_tokens": 14158834.0,
      "step": 6175
    },
    {
      "entropy": 5.415588140487671,
      "epoch": 0.5936599423631124,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004972035353609923,
      "loss": 5.2878,
      "mean_token_accuracy": 0.19879126250743867,
      "num_tokens": 14170694.0,
      "step": 6180
    },
    {
      "entropy": 5.386823749542236,
      "epoch": 0.5941402497598463,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004971981443717572,
      "loss": 5.1986,
      "mean_token_accuracy": 0.20433641821146012,
      "num_tokens": 14183184.0,
      "step": 6185
    },
    {
      "entropy": 5.507526016235351,
      "epoch": 0.5946205571565802,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004971927482237409,
      "loss": 5.2783,
      "mean_token_accuracy": 0.1958542063832283,
      "num_tokens": 14194761.0,
      "step": 6190
    },
    {
      "entropy": 5.362305402755737,
      "epoch": 0.5951008645533141,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004971873469170689,
      "loss": 5.0956,
      "mean_token_accuracy": 0.20270321220159532,
      "num_tokens": 14205820.0,
      "step": 6195
    },
    {
      "entropy": 5.2876345157623295,
      "epoch": 0.595581171950048,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004971819404518664,
      "loss": 5.1111,
      "mean_token_accuracy": 0.212001796066761,
      "num_tokens": 14217826.0,
      "step": 6200
    },
    {
      "entropy": 5.361300897598267,
      "epoch": 0.5960614793467819,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000497176528828259,
      "loss": 5.0835,
      "mean_token_accuracy": 0.21398791074752807,
      "num_tokens": 14228632.0,
      "step": 6205
    },
    {
      "entropy": 5.4559839248657225,
      "epoch": 0.5965417867435159,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004971711120463722,
      "loss": 5.3041,
      "mean_token_accuracy": 0.1970059245824814,
      "num_tokens": 14240231.0,
      "step": 6210
    },
    {
      "entropy": 5.486827707290649,
      "epoch": 0.5970220941402498,
      "grad_norm": 0.96875,
      "learning_rate": 0.000497165690106332,
      "loss": 5.1409,
      "mean_token_accuracy": 0.20812199562788009,
      "num_tokens": 14251718.0,
      "step": 6215
    },
    {
      "entropy": 5.3630602836608885,
      "epoch": 0.5975024015369836,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004971602630082642,
      "loss": 5.1945,
      "mean_token_accuracy": 0.2097410961985588,
      "num_tokens": 14263129.0,
      "step": 6220
    },
    {
      "entropy": 5.300925445556641,
      "epoch": 0.5979827089337176,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004971548307522947,
      "loss": 5.1044,
      "mean_token_accuracy": 0.20420224219560623,
      "num_tokens": 14274465.0,
      "step": 6225
    },
    {
      "entropy": 5.499400806427002,
      "epoch": 0.5984630163304515,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004971493933385498,
      "loss": 5.2948,
      "mean_token_accuracy": 0.20205555707216263,
      "num_tokens": 14286190.0,
      "step": 6230
    },
    {
      "entropy": 5.444535160064698,
      "epoch": 0.5989433237271854,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004971439507671556,
      "loss": 5.2302,
      "mean_token_accuracy": 0.20107307583093642,
      "num_tokens": 14297638.0,
      "step": 6235
    },
    {
      "entropy": 5.361764144897461,
      "epoch": 0.5994236311239193,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004971385030382384,
      "loss": 5.1709,
      "mean_token_accuracy": 0.20606155544519425,
      "num_tokens": 14309773.0,
      "step": 6240
    },
    {
      "entropy": 5.37435564994812,
      "epoch": 0.5999039385206533,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004971330501519248,
      "loss": 5.1555,
      "mean_token_accuracy": 0.2098422646522522,
      "num_tokens": 14320543.0,
      "step": 6245
    },
    {
      "entropy": 5.445180654525757,
      "epoch": 0.6003842459173871,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004971275921083414,
      "loss": 5.2277,
      "mean_token_accuracy": 0.19865779876708983,
      "num_tokens": 14332076.0,
      "step": 6250
    },
    {
      "entropy": 5.333190631866455,
      "epoch": 0.600864553314121,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000497122128907615,
      "loss": 5.1425,
      "mean_token_accuracy": 0.20159315317869186,
      "num_tokens": 14343768.0,
      "step": 6255
    },
    {
      "entropy": 5.309207582473755,
      "epoch": 0.6013448607108549,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004971166605498722,
      "loss": 5.1857,
      "mean_token_accuracy": 0.19793325811624526,
      "num_tokens": 14357051.0,
      "step": 6260
    },
    {
      "entropy": 5.4213268756866455,
      "epoch": 0.6018251681075889,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004971111870352401,
      "loss": 5.2038,
      "mean_token_accuracy": 0.20178428441286086,
      "num_tokens": 14368800.0,
      "step": 6265
    },
    {
      "entropy": 5.413872814178466,
      "epoch": 0.6023054755043228,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004971057083638458,
      "loss": 5.181,
      "mean_token_accuracy": 0.20945534408092498,
      "num_tokens": 14379617.0,
      "step": 6270
    },
    {
      "entropy": 5.3219886302948,
      "epoch": 0.6027857829010567,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004971002245358166,
      "loss": 5.1684,
      "mean_token_accuracy": 0.20470429807901383,
      "num_tokens": 14391454.0,
      "step": 6275
    },
    {
      "entropy": 5.4138518333435055,
      "epoch": 0.6032660902977905,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970947355512795,
      "loss": 5.1666,
      "mean_token_accuracy": 0.20715542882680893,
      "num_tokens": 14402379.0,
      "step": 6280
    },
    {
      "entropy": 5.361328744888306,
      "epoch": 0.6037463976945245,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004970892414103622,
      "loss": 5.1526,
      "mean_token_accuracy": 0.20690900087356567,
      "num_tokens": 14415040.0,
      "step": 6285
    },
    {
      "entropy": 5.355616760253906,
      "epoch": 0.6042267050912584,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970837421131921,
      "loss": 5.2586,
      "mean_token_accuracy": 0.1893886595964432,
      "num_tokens": 14426677.0,
      "step": 6290
    },
    {
      "entropy": 5.423394632339478,
      "epoch": 0.6047070124879923,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004970782376598972,
      "loss": 5.1627,
      "mean_token_accuracy": 0.20687107294797896,
      "num_tokens": 14436676.0,
      "step": 6295
    },
    {
      "entropy": 5.3968834400177,
      "epoch": 0.6051873198847262,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004970727280506048,
      "loss": 5.1978,
      "mean_token_accuracy": 0.19833542108535768,
      "num_tokens": 14448294.0,
      "step": 6300
    },
    {
      "entropy": 5.334224081039428,
      "epoch": 0.6056676272814602,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970672132854431,
      "loss": 5.1789,
      "mean_token_accuracy": 0.202311372756958,
      "num_tokens": 14460642.0,
      "step": 6305
    },
    {
      "entropy": 5.3564201354980465,
      "epoch": 0.6061479346781941,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004970616933645403,
      "loss": 5.0875,
      "mean_token_accuracy": 0.21412288546562194,
      "num_tokens": 14471370.0,
      "step": 6310
    },
    {
      "entropy": 5.464383506774903,
      "epoch": 0.6066282420749279,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004970561682880242,
      "loss": 5.2479,
      "mean_token_accuracy": 0.2040972739458084,
      "num_tokens": 14482358.0,
      "step": 6315
    },
    {
      "entropy": 5.392704248428345,
      "epoch": 0.6071085494716618,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004970506380560233,
      "loss": 5.262,
      "mean_token_accuracy": 0.20550439059734343,
      "num_tokens": 14494413.0,
      "step": 6320
    },
    {
      "entropy": 5.447866010665893,
      "epoch": 0.6075888568683958,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004970451026686659,
      "loss": 5.1857,
      "mean_token_accuracy": 0.210190112888813,
      "num_tokens": 14506370.0,
      "step": 6325
    },
    {
      "entropy": 5.41877064704895,
      "epoch": 0.6080691642651297,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004970395621260806,
      "loss": 5.1776,
      "mean_token_accuracy": 0.20847591310739516,
      "num_tokens": 14517198.0,
      "step": 6330
    },
    {
      "entropy": 5.458759832382202,
      "epoch": 0.6085494716618636,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497034016428396,
      "loss": 5.271,
      "mean_token_accuracy": 0.19544457793235778,
      "num_tokens": 14529434.0,
      "step": 6335
    },
    {
      "entropy": 5.3051248550415036,
      "epoch": 0.6090297790585975,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004970284655757409,
      "loss": 5.1659,
      "mean_token_accuracy": 0.2054130107164383,
      "num_tokens": 14540857.0,
      "step": 6340
    },
    {
      "entropy": 5.3334784507751465,
      "epoch": 0.6095100864553314,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004970229095682439,
      "loss": 5.1763,
      "mean_token_accuracy": 0.20291784554719924,
      "num_tokens": 14552594.0,
      "step": 6345
    },
    {
      "entropy": 5.425951862335205,
      "epoch": 0.6099903938520653,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970173484060344,
      "loss": 5.2429,
      "mean_token_accuracy": 0.1976109728217125,
      "num_tokens": 14563599.0,
      "step": 6350
    },
    {
      "entropy": 5.430559301376343,
      "epoch": 0.6104707012487992,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004970117820892414,
      "loss": 5.183,
      "mean_token_accuracy": 0.20517935752868652,
      "num_tokens": 14575905.0,
      "step": 6355
    },
    {
      "entropy": 5.439344120025635,
      "epoch": 0.6109510086455331,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970062106179939,
      "loss": 5.2138,
      "mean_token_accuracy": 0.21136917918920517,
      "num_tokens": 14587800.0,
      "step": 6360
    },
    {
      "entropy": 5.378159379959106,
      "epoch": 0.6114313160422671,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004970006339924214,
      "loss": 5.1379,
      "mean_token_accuracy": 0.20937149822711945,
      "num_tokens": 14600654.0,
      "step": 6365
    },
    {
      "entropy": 5.400775575637818,
      "epoch": 0.611911623439001,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004969950522126534,
      "loss": 5.1397,
      "mean_token_accuracy": 0.2076369881629944,
      "num_tokens": 14611985.0,
      "step": 6370
    },
    {
      "entropy": 5.348199558258057,
      "epoch": 0.6123919308357348,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004969894652788196,
      "loss": 5.1455,
      "mean_token_accuracy": 0.20937911719083785,
      "num_tokens": 14625004.0,
      "step": 6375
    },
    {
      "entropy": 5.408209419250488,
      "epoch": 0.6128722382324687,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004969838731910494,
      "loss": 5.2091,
      "mean_token_accuracy": 0.20719991624355316,
      "num_tokens": 14635381.0,
      "step": 6380
    },
    {
      "entropy": 5.332770204544067,
      "epoch": 0.6133525456292027,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969782759494729,
      "loss": 5.1532,
      "mean_token_accuracy": 0.2092129647731781,
      "num_tokens": 14646582.0,
      "step": 6385
    },
    {
      "entropy": 5.3539073944091795,
      "epoch": 0.6138328530259366,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00049697267355422,
      "loss": 5.2194,
      "mean_token_accuracy": 0.20976134836673738,
      "num_tokens": 14658024.0,
      "step": 6390
    },
    {
      "entropy": 5.5265398025512695,
      "epoch": 0.6143131604226705,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004969670660054208,
      "loss": 5.2474,
      "mean_token_accuracy": 0.1984589397907257,
      "num_tokens": 14669933.0,
      "step": 6395
    },
    {
      "entropy": 5.395998096466064,
      "epoch": 0.6147934678194045,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004969614533032054,
      "loss": 5.1977,
      "mean_token_accuracy": 0.20416766554117202,
      "num_tokens": 14681331.0,
      "step": 6400
    },
    {
      "entropy": 5.308030843734741,
      "epoch": 0.6152737752161384,
      "grad_norm": 0.875,
      "learning_rate": 0.0004969558354477041,
      "loss": 5.0594,
      "mean_token_accuracy": 0.21813455373048782,
      "num_tokens": 14691361.0,
      "step": 6405
    },
    {
      "entropy": 5.355904817581177,
      "epoch": 0.6157540826128722,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004969502124390474,
      "loss": 5.151,
      "mean_token_accuracy": 0.20957321375608445,
      "num_tokens": 14701791.0,
      "step": 6410
    },
    {
      "entropy": 5.40302791595459,
      "epoch": 0.6162343900096061,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969445842773658,
      "loss": 5.2122,
      "mean_token_accuracy": 0.1992111697793007,
      "num_tokens": 14713672.0,
      "step": 6415
    },
    {
      "entropy": 5.364359617233276,
      "epoch": 0.6167146974063401,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049693895096279,
      "loss": 5.0629,
      "mean_token_accuracy": 0.21054122298955918,
      "num_tokens": 14724789.0,
      "step": 6420
    },
    {
      "entropy": 5.355506849288941,
      "epoch": 0.617195004803074,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004969333124954508,
      "loss": 5.1385,
      "mean_token_accuracy": 0.20256014466285704,
      "num_tokens": 14737212.0,
      "step": 6425
    },
    {
      "entropy": 5.331637668609619,
      "epoch": 0.6176753121998079,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004969276688754791,
      "loss": 5.1553,
      "mean_token_accuracy": 0.20373374074697495,
      "num_tokens": 14748387.0,
      "step": 6430
    },
    {
      "entropy": 5.381811285018921,
      "epoch": 0.6181556195965417,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004969220201030059,
      "loss": 5.1926,
      "mean_token_accuracy": 0.20251691937446595,
      "num_tokens": 14758477.0,
      "step": 6435
    },
    {
      "entropy": 5.4244917869567875,
      "epoch": 0.6186359269932757,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004969163661781624,
      "loss": 5.2765,
      "mean_token_accuracy": 0.20467588901519776,
      "num_tokens": 14769650.0,
      "step": 6440
    },
    {
      "entropy": 5.304611587524414,
      "epoch": 0.6191162343900096,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004969107071010798,
      "loss": 5.0545,
      "mean_token_accuracy": 0.21091441065073013,
      "num_tokens": 14780988.0,
      "step": 6445
    },
    {
      "entropy": 5.3784322261810305,
      "epoch": 0.6195965417867435,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004969050428718895,
      "loss": 5.1588,
      "mean_token_accuracy": 0.2065223604440689,
      "num_tokens": 14792458.0,
      "step": 6450
    },
    {
      "entropy": 5.419982624053955,
      "epoch": 0.6200768491834774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000496899373490723,
      "loss": 5.233,
      "mean_token_accuracy": 0.2050531879067421,
      "num_tokens": 14804772.0,
      "step": 6455
    },
    {
      "entropy": 5.415715646743775,
      "epoch": 0.6205571565802114,
      "grad_norm": 0.859375,
      "learning_rate": 0.000496893698957712,
      "loss": 5.1652,
      "mean_token_accuracy": 0.20256629437208176,
      "num_tokens": 14816126.0,
      "step": 6460
    },
    {
      "entropy": 5.417246246337891,
      "epoch": 0.6210374639769453,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004968880192729882,
      "loss": 5.2784,
      "mean_token_accuracy": 0.19865417778491973,
      "num_tokens": 14829376.0,
      "step": 6465
    },
    {
      "entropy": 5.343904733657837,
      "epoch": 0.6215177713736791,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004968823344366835,
      "loss": 5.0871,
      "mean_token_accuracy": 0.2137269049882889,
      "num_tokens": 14841098.0,
      "step": 6470
    },
    {
      "entropy": 5.354584169387818,
      "epoch": 0.621998078770413,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004968766444489298,
      "loss": 5.1264,
      "mean_token_accuracy": 0.20900900065898895,
      "num_tokens": 14852690.0,
      "step": 6475
    },
    {
      "entropy": 5.379853916168213,
      "epoch": 0.622478386167147,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004968709493098593,
      "loss": 5.1067,
      "mean_token_accuracy": 0.21387286633253097,
      "num_tokens": 14863327.0,
      "step": 6480
    },
    {
      "entropy": 5.446810865402222,
      "epoch": 0.6229586935638809,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004968652490196041,
      "loss": 5.2832,
      "mean_token_accuracy": 0.20186906754970552,
      "num_tokens": 14875213.0,
      "step": 6485
    },
    {
      "entropy": 5.405354928970337,
      "epoch": 0.6234390009606148,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004968595435782967,
      "loss": 5.1779,
      "mean_token_accuracy": 0.21098628044128417,
      "num_tokens": 14886129.0,
      "step": 6490
    },
    {
      "entropy": 5.459101867675781,
      "epoch": 0.6239193083573487,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004968538329860695,
      "loss": 5.3211,
      "mean_token_accuracy": 0.19605984836816787,
      "num_tokens": 14897217.0,
      "step": 6495
    },
    {
      "entropy": 5.4209792137146,
      "epoch": 0.6243996157540826,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004968481172430549,
      "loss": 5.2365,
      "mean_token_accuracy": 0.2023770898580551,
      "num_tokens": 14908438.0,
      "step": 6500
    },
    {
      "entropy": 5.363001203536987,
      "epoch": 0.6248799231508165,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000496842396349386,
      "loss": 5.1093,
      "mean_token_accuracy": 0.21454905420541764,
      "num_tokens": 14920472.0,
      "step": 6505
    },
    {
      "entropy": 5.318320798873901,
      "epoch": 0.6253602305475504,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968366703051952,
      "loss": 5.1638,
      "mean_token_accuracy": 0.20634848922491072,
      "num_tokens": 14932075.0,
      "step": 6510
    },
    {
      "entropy": 5.345568227767944,
      "epoch": 0.6258405379442843,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004968309391106157,
      "loss": 5.2011,
      "mean_token_accuracy": 0.20254727452993393,
      "num_tokens": 14942971.0,
      "step": 6515
    },
    {
      "entropy": 5.369955396652221,
      "epoch": 0.6263208453410183,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004968252027657806,
      "loss": 5.0396,
      "mean_token_accuracy": 0.21421653479337693,
      "num_tokens": 14954288.0,
      "step": 6520
    },
    {
      "entropy": 5.288963842391968,
      "epoch": 0.6268011527377522,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004968194612708229,
      "loss": 5.1158,
      "mean_token_accuracy": 0.2066323310136795,
      "num_tokens": 14966017.0,
      "step": 6525
    },
    {
      "entropy": 5.424745082855225,
      "epoch": 0.627281460134486,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004968137146258759,
      "loss": 5.1844,
      "mean_token_accuracy": 0.20647535622119903,
      "num_tokens": 14978022.0,
      "step": 6530
    },
    {
      "entropy": 5.30846676826477,
      "epoch": 0.6277617675312199,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004968079628310732,
      "loss": 5.0942,
      "mean_token_accuracy": 0.21247957795858383,
      "num_tokens": 14990370.0,
      "step": 6535
    },
    {
      "entropy": 5.300876474380493,
      "epoch": 0.6282420749279539,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004968022058865482,
      "loss": 5.1071,
      "mean_token_accuracy": 0.21693088710308076,
      "num_tokens": 15001535.0,
      "step": 6540
    },
    {
      "entropy": 5.42994966506958,
      "epoch": 0.6287223823246878,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004967964437924346,
      "loss": 5.2046,
      "mean_token_accuracy": 0.20321286767721175,
      "num_tokens": 15012560.0,
      "step": 6545
    },
    {
      "entropy": 5.2918657779693605,
      "epoch": 0.6292026897214217,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004967906765488662,
      "loss": 5.1063,
      "mean_token_accuracy": 0.2127204954624176,
      "num_tokens": 15024594.0,
      "step": 6550
    },
    {
      "entropy": 5.364038038253784,
      "epoch": 0.6296829971181557,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004967849041559769,
      "loss": 5.123,
      "mean_token_accuracy": 0.20624317228794098,
      "num_tokens": 15034735.0,
      "step": 6555
    },
    {
      "entropy": 5.428404045104981,
      "epoch": 0.6301633045148896,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004967791266139006,
      "loss": 5.2345,
      "mean_token_accuracy": 0.20093282610177993,
      "num_tokens": 15046955.0,
      "step": 6560
    },
    {
      "entropy": 5.367275381088257,
      "epoch": 0.6306436119116234,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004967733439227716,
      "loss": 5.1069,
      "mean_token_accuracy": 0.20834526866674424,
      "num_tokens": 15058250.0,
      "step": 6565
    },
    {
      "entropy": 5.391459703445435,
      "epoch": 0.6311239193083573,
      "grad_norm": 0.859375,
      "learning_rate": 0.000496767556082724,
      "loss": 5.2314,
      "mean_token_accuracy": 0.20451269298791885,
      "num_tokens": 15069451.0,
      "step": 6570
    },
    {
      "entropy": 5.371989822387695,
      "epoch": 0.6316042267050913,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004967617630938924,
      "loss": 5.1982,
      "mean_token_accuracy": 0.2065337836742401,
      "num_tokens": 15080920.0,
      "step": 6575
    },
    {
      "entropy": 5.330305576324463,
      "epoch": 0.6320845341018252,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000496755964956411,
      "loss": 5.1283,
      "mean_token_accuracy": 0.20611016005277633,
      "num_tokens": 15091579.0,
      "step": 6580
    },
    {
      "entropy": 5.382533836364746,
      "epoch": 0.6325648414985591,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004967501616704147,
      "loss": 5.23,
      "mean_token_accuracy": 0.20330362915992736,
      "num_tokens": 15103097.0,
      "step": 6585
    },
    {
      "entropy": 5.359675598144531,
      "epoch": 0.633045148895293,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000496744353236038,
      "loss": 5.1169,
      "mean_token_accuracy": 0.20733192414045334,
      "num_tokens": 15114646.0,
      "step": 6590
    },
    {
      "entropy": 5.445838499069214,
      "epoch": 0.633525456292027,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000496738539653416,
      "loss": 5.1716,
      "mean_token_accuracy": 0.20627779513597488,
      "num_tokens": 15126019.0,
      "step": 6595
    },
    {
      "entropy": 5.388510131835938,
      "epoch": 0.6340057636887608,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004967327209226835,
      "loss": 5.1574,
      "mean_token_accuracy": 0.21125543713569642,
      "num_tokens": 15136132.0,
      "step": 6600
    },
    {
      "entropy": 5.361651802062989,
      "epoch": 0.6344860710854947,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004967268970439755,
      "loss": 5.2463,
      "mean_token_accuracy": 0.20033118426799773,
      "num_tokens": 15149016.0,
      "step": 6605
    },
    {
      "entropy": 5.43954291343689,
      "epoch": 0.6349663784822286,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004967210680174274,
      "loss": 5.2067,
      "mean_token_accuracy": 0.21057035475969316,
      "num_tokens": 15160302.0,
      "step": 6610
    },
    {
      "entropy": 5.406799554824829,
      "epoch": 0.6354466858789626,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004967152338431745,
      "loss": 5.1898,
      "mean_token_accuracy": 0.20454303473234176,
      "num_tokens": 15171770.0,
      "step": 6615
    },
    {
      "entropy": 5.354743099212646,
      "epoch": 0.6359269932756965,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967093945213523,
      "loss": 5.0746,
      "mean_token_accuracy": 0.20936929285526276,
      "num_tokens": 15182283.0,
      "step": 6620
    },
    {
      "entropy": 5.300932168960571,
      "epoch": 0.6364073006724303,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004967035500520962,
      "loss": 5.1928,
      "mean_token_accuracy": 0.20960211157798767,
      "num_tokens": 15193917.0,
      "step": 6625
    },
    {
      "entropy": 5.48442587852478,
      "epoch": 0.6368876080691642,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004966977004355421,
      "loss": 5.2463,
      "mean_token_accuracy": 0.1999865621328354,
      "num_tokens": 15205854.0,
      "step": 6630
    },
    {
      "entropy": 5.303269624710083,
      "epoch": 0.6373679154658982,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004966918456718256,
      "loss": 5.1233,
      "mean_token_accuracy": 0.20779794603586196,
      "num_tokens": 15217884.0,
      "step": 6635
    },
    {
      "entropy": 5.403727245330811,
      "epoch": 0.6378482228626321,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966859857610828,
      "loss": 5.248,
      "mean_token_accuracy": 0.20089327096939086,
      "num_tokens": 15229493.0,
      "step": 6640
    },
    {
      "entropy": 5.373949956893921,
      "epoch": 0.638328530259366,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004966801207034497,
      "loss": 5.1924,
      "mean_token_accuracy": 0.1991853430867195,
      "num_tokens": 15241402.0,
      "step": 6645
    },
    {
      "entropy": 5.421070337295532,
      "epoch": 0.6388088376560999,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004966742504990624,
      "loss": 5.2059,
      "mean_token_accuracy": 0.20060463845729828,
      "num_tokens": 15252981.0,
      "step": 6650
    },
    {
      "entropy": 5.390252637863159,
      "epoch": 0.6392891450528339,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004966683751480572,
      "loss": 5.1317,
      "mean_token_accuracy": 0.21188561022281646,
      "num_tokens": 15264171.0,
      "step": 6655
    },
    {
      "entropy": 5.336518430709839,
      "epoch": 0.6397694524495677,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004966624946505706,
      "loss": 5.1287,
      "mean_token_accuracy": 0.21237500607967377,
      "num_tokens": 15275595.0,
      "step": 6660
    },
    {
      "entropy": 5.411849784851074,
      "epoch": 0.6402497598463016,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004966566090067391,
      "loss": 5.2162,
      "mean_token_accuracy": 0.20244170725345612,
      "num_tokens": 15286074.0,
      "step": 6665
    },
    {
      "entropy": 5.4499588966369625,
      "epoch": 0.6407300672430355,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004966507182166993,
      "loss": 5.1739,
      "mean_token_accuracy": 0.20561551451683044,
      "num_tokens": 15296230.0,
      "step": 6670
    },
    {
      "entropy": 5.350120973587036,
      "epoch": 0.6412103746397695,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000496644822280588,
      "loss": 5.1463,
      "mean_token_accuracy": 0.2097763329744339,
      "num_tokens": 15307428.0,
      "step": 6675
    },
    {
      "entropy": 5.427899646759033,
      "epoch": 0.6416906820365034,
      "grad_norm": 0.890625,
      "learning_rate": 0.000496638921198542,
      "loss": 5.2659,
      "mean_token_accuracy": 0.19720300137996674,
      "num_tokens": 15318796.0,
      "step": 6680
    },
    {
      "entropy": 5.370028114318847,
      "epoch": 0.6421709894332372,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004966330149706984,
      "loss": 5.1689,
      "mean_token_accuracy": 0.20950590670108796,
      "num_tokens": 15330914.0,
      "step": 6685
    },
    {
      "entropy": 5.358871841430664,
      "epoch": 0.6426512968299711,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004966271035971944,
      "loss": 5.1317,
      "mean_token_accuracy": 0.20505664646625518,
      "num_tokens": 15341756.0,
      "step": 6690
    },
    {
      "entropy": 5.40494704246521,
      "epoch": 0.6431316042267051,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004966211870781671,
      "loss": 5.2342,
      "mean_token_accuracy": 0.20485753268003465,
      "num_tokens": 15353217.0,
      "step": 6695
    },
    {
      "entropy": 5.4200887203216555,
      "epoch": 0.643611911623439,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000496615265413754,
      "loss": 5.1282,
      "mean_token_accuracy": 0.19946543574333192,
      "num_tokens": 15363813.0,
      "step": 6700
    },
    {
      "entropy": 5.352343225479126,
      "epoch": 0.6440922190201729,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966093386040923,
      "loss": 5.1453,
      "mean_token_accuracy": 0.20322655290365219,
      "num_tokens": 15375223.0,
      "step": 6705
    },
    {
      "entropy": 5.3955957889556885,
      "epoch": 0.6445725264169068,
      "grad_norm": 1.0,
      "learning_rate": 0.00049660340664932,
      "loss": 5.1331,
      "mean_token_accuracy": 0.20524256229400634,
      "num_tokens": 15385523.0,
      "step": 6710
    },
    {
      "entropy": 5.408837413787841,
      "epoch": 0.6450528338136408,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004965974695495746,
      "loss": 5.1991,
      "mean_token_accuracy": 0.20230422168970108,
      "num_tokens": 15397262.0,
      "step": 6715
    },
    {
      "entropy": 5.305728101730347,
      "epoch": 0.6455331412103746,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004965915273049938,
      "loss": 5.1204,
      "mean_token_accuracy": 0.20638727247714997,
      "num_tokens": 15409043.0,
      "step": 6720
    },
    {
      "entropy": 5.378109169006348,
      "epoch": 0.6460134486071085,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004965855799157158,
      "loss": 5.1304,
      "mean_token_accuracy": 0.2075420618057251,
      "num_tokens": 15420859.0,
      "step": 6725
    },
    {
      "entropy": 5.3898787021636965,
      "epoch": 0.6464937560038425,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004965796273818787,
      "loss": 5.2029,
      "mean_token_accuracy": 0.20161716938018798,
      "num_tokens": 15433650.0,
      "step": 6730
    },
    {
      "entropy": 5.327402925491333,
      "epoch": 0.6469740634005764,
      "grad_norm": 1.0,
      "learning_rate": 0.0004965736697036206,
      "loss": 5.0872,
      "mean_token_accuracy": 0.21341919004917145,
      "num_tokens": 15445158.0,
      "step": 6735
    },
    {
      "entropy": 5.426046419143677,
      "epoch": 0.6474543707973103,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004965677068810798,
      "loss": 5.3573,
      "mean_token_accuracy": 0.19311742782592772,
      "num_tokens": 15456974.0,
      "step": 6740
    },
    {
      "entropy": 5.446528100967408,
      "epoch": 0.6479346781940442,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004965617389143947,
      "loss": 5.2157,
      "mean_token_accuracy": 0.20665405094623565,
      "num_tokens": 15467395.0,
      "step": 6745
    },
    {
      "entropy": 5.346460294723511,
      "epoch": 0.6484149855907781,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000496555765803704,
      "loss": 5.2038,
      "mean_token_accuracy": 0.20146800428628922,
      "num_tokens": 15479396.0,
      "step": 6750
    },
    {
      "entropy": 5.360586452484131,
      "epoch": 0.648895292987512,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004965497875491462,
      "loss": 5.1588,
      "mean_token_accuracy": 0.20793106108903886,
      "num_tokens": 15490992.0,
      "step": 6755
    },
    {
      "entropy": 5.511486053466797,
      "epoch": 0.6493756003842459,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004965438041508604,
      "loss": 5.2276,
      "mean_token_accuracy": 0.19770817160606385,
      "num_tokens": 15502413.0,
      "step": 6760
    },
    {
      "entropy": 5.392805480957032,
      "epoch": 0.6498559077809798,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004965378156089851,
      "loss": 5.1637,
      "mean_token_accuracy": 0.19902628809213638,
      "num_tokens": 15512588.0,
      "step": 6765
    },
    {
      "entropy": 5.365663719177246,
      "epoch": 0.6503362151777138,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004965318219236597,
      "loss": 5.1508,
      "mean_token_accuracy": 0.2032195508480072,
      "num_tokens": 15523252.0,
      "step": 6770
    },
    {
      "entropy": 5.322205829620361,
      "epoch": 0.6508165225744477,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004965258230950232,
      "loss": 5.2475,
      "mean_token_accuracy": 0.20265621691942215,
      "num_tokens": 15534975.0,
      "step": 6775
    },
    {
      "entropy": 5.382902050018311,
      "epoch": 0.6512968299711815,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004965198191232148,
      "loss": 5.1091,
      "mean_token_accuracy": 0.207707180082798,
      "num_tokens": 15545709.0,
      "step": 6780
    },
    {
      "entropy": 5.433700752258301,
      "epoch": 0.6517771373679154,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496513810008374,
      "loss": 5.2043,
      "mean_token_accuracy": 0.20900945365428925,
      "num_tokens": 15557270.0,
      "step": 6785
    },
    {
      "entropy": 5.403327751159668,
      "epoch": 0.6522574447646494,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004965077957506403,
      "loss": 5.1715,
      "mean_token_accuracy": 0.20481457859277724,
      "num_tokens": 15569156.0,
      "step": 6790
    },
    {
      "entropy": 5.327019262313843,
      "epoch": 0.6527377521613833,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004965017763501533,
      "loss": 5.1101,
      "mean_token_accuracy": 0.20782879143953323,
      "num_tokens": 15579270.0,
      "step": 6795
    },
    {
      "entropy": 5.294063758850098,
      "epoch": 0.6532180595581172,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004964957518070529,
      "loss": 5.0712,
      "mean_token_accuracy": 0.2134845018386841,
      "num_tokens": 15589912.0,
      "step": 6800
    },
    {
      "entropy": 5.3868248462677,
      "epoch": 0.6536983669548511,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004964897221214788,
      "loss": 5.1587,
      "mean_token_accuracy": 0.20910485982894897,
      "num_tokens": 15601088.0,
      "step": 6805
    },
    {
      "entropy": 5.455225038528442,
      "epoch": 0.654178674351585,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000496483687293571,
      "loss": 5.1886,
      "mean_token_accuracy": 0.2052254095673561,
      "num_tokens": 15612520.0,
      "step": 6810
    },
    {
      "entropy": 5.349751377105713,
      "epoch": 0.6546589817483189,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004964776473234696,
      "loss": 5.0978,
      "mean_token_accuracy": 0.21420973986387254,
      "num_tokens": 15623991.0,
      "step": 6815
    },
    {
      "entropy": 5.252722883224488,
      "epoch": 0.6551392891450528,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000496471602211315,
      "loss": 5.0387,
      "mean_token_accuracy": 0.2098737835884094,
      "num_tokens": 15635389.0,
      "step": 6820
    },
    {
      "entropy": 5.355444765090942,
      "epoch": 0.6556195965417867,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004964655519572475,
      "loss": 5.1251,
      "mean_token_accuracy": 0.2136618196964264,
      "num_tokens": 15646427.0,
      "step": 6825
    },
    {
      "entropy": 5.259690999984741,
      "epoch": 0.6560999039385207,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004964594965614072,
      "loss": 5.1466,
      "mean_token_accuracy": 0.20416057854890823,
      "num_tokens": 15657518.0,
      "step": 6830
    },
    {
      "entropy": 5.414389944076538,
      "epoch": 0.6565802113352546,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004964534360239353,
      "loss": 5.1564,
      "mean_token_accuracy": 0.20021041184663774,
      "num_tokens": 15669775.0,
      "step": 6835
    },
    {
      "entropy": 5.365410375595093,
      "epoch": 0.6570605187319885,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000496447370344972,
      "loss": 5.1458,
      "mean_token_accuracy": 0.20207538455724716,
      "num_tokens": 15682566.0,
      "step": 6840
    },
    {
      "entropy": 5.304757499694825,
      "epoch": 0.6575408261287223,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004964412995246584,
      "loss": 5.1465,
      "mean_token_accuracy": 0.21071666330099106,
      "num_tokens": 15693168.0,
      "step": 6845
    },
    {
      "entropy": 5.376939153671264,
      "epoch": 0.6580211335254563,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004964352235631354,
      "loss": 5.1636,
      "mean_token_accuracy": 0.2066321477293968,
      "num_tokens": 15703879.0,
      "step": 6850
    },
    {
      "entropy": 5.442447233200073,
      "epoch": 0.6585014409221902,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000496429142460544,
      "loss": 5.1921,
      "mean_token_accuracy": 0.1971552312374115,
      "num_tokens": 15716015.0,
      "step": 6855
    },
    {
      "entropy": 5.444846773147583,
      "epoch": 0.6589817483189241,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004964230562170254,
      "loss": 5.212,
      "mean_token_accuracy": 0.20607330352067948,
      "num_tokens": 15728254.0,
      "step": 6860
    },
    {
      "entropy": 5.373237371444702,
      "epoch": 0.659462055715658,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004964169648327209,
      "loss": 5.1668,
      "mean_token_accuracy": 0.2062647521495819,
      "num_tokens": 15738778.0,
      "step": 6865
    },
    {
      "entropy": 5.4141106605529785,
      "epoch": 0.659942363112392,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000496410868307772,
      "loss": 5.1785,
      "mean_token_accuracy": 0.20512595921754836,
      "num_tokens": 15750305.0,
      "step": 6870
    },
    {
      "entropy": 5.410314226150513,
      "epoch": 0.6604226705091258,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964047666423203,
      "loss": 5.1535,
      "mean_token_accuracy": 0.20569400489330292,
      "num_tokens": 15761303.0,
      "step": 6875
    },
    {
      "entropy": 5.319344091415405,
      "epoch": 0.6609029779058597,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004963986598365072,
      "loss": 5.1133,
      "mean_token_accuracy": 0.21337104141712188,
      "num_tokens": 15773095.0,
      "step": 6880
    },
    {
      "entropy": 5.391595458984375,
      "epoch": 0.6613832853025937,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963925478904746,
      "loss": 5.1813,
      "mean_token_accuracy": 0.21176904290914536,
      "num_tokens": 15784405.0,
      "step": 6885
    },
    {
      "entropy": 5.264011859893799,
      "epoch": 0.6618635926993276,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004963864308043645,
      "loss": 5.1088,
      "mean_token_accuracy": 0.2080717846751213,
      "num_tokens": 15795178.0,
      "step": 6890
    },
    {
      "entropy": 5.348097467422486,
      "epoch": 0.6623439000960615,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004963803085783189,
      "loss": 5.1305,
      "mean_token_accuracy": 0.207475383579731,
      "num_tokens": 15806205.0,
      "step": 6895
    },
    {
      "entropy": 5.346061849594117,
      "epoch": 0.6628242074927954,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004963741812124799,
      "loss": 5.0563,
      "mean_token_accuracy": 0.2159065529704094,
      "num_tokens": 15817474.0,
      "step": 6900
    },
    {
      "entropy": 5.296646690368652,
      "epoch": 0.6633045148895294,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004963680487069898,
      "loss": 5.1637,
      "mean_token_accuracy": 0.20098544657230377,
      "num_tokens": 15829728.0,
      "step": 6905
    },
    {
      "entropy": 5.423515939712525,
      "epoch": 0.6637848222862632,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004963619110619908,
      "loss": 5.2504,
      "mean_token_accuracy": 0.2025123044848442,
      "num_tokens": 15840082.0,
      "step": 6910
    },
    {
      "entropy": 5.407306671142578,
      "epoch": 0.6642651296829971,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963557682776256,
      "loss": 5.126,
      "mean_token_accuracy": 0.20657253712415696,
      "num_tokens": 15851450.0,
      "step": 6915
    },
    {
      "entropy": 5.413578653335572,
      "epoch": 0.664745437079731,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004963496203540368,
      "loss": 5.2877,
      "mean_token_accuracy": 0.19798053354024886,
      "num_tokens": 15864168.0,
      "step": 6920
    },
    {
      "entropy": 5.421274948120117,
      "epoch": 0.665225744476465,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004963434672913671,
      "loss": 5.2048,
      "mean_token_accuracy": 0.2082076609134674,
      "num_tokens": 15875634.0,
      "step": 6925
    },
    {
      "entropy": 5.465573310852051,
      "epoch": 0.6657060518731989,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004963373090897592,
      "loss": 5.2337,
      "mean_token_accuracy": 0.20044171810150146,
      "num_tokens": 15888411.0,
      "step": 6930
    },
    {
      "entropy": 5.374470853805542,
      "epoch": 0.6661863592699327,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004963311457493563,
      "loss": 5.2215,
      "mean_token_accuracy": 0.20023801028728486,
      "num_tokens": 15901084.0,
      "step": 6935
    },
    {
      "entropy": 5.479472970962524,
      "epoch": 0.6666666666666666,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004963249772703015,
      "loss": 5.2402,
      "mean_token_accuracy": 0.1996363654732704,
      "num_tokens": 15912061.0,
      "step": 6940
    },
    {
      "entropy": 5.437343740463257,
      "epoch": 0.6671469740634006,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004963188036527378,
      "loss": 5.2838,
      "mean_token_accuracy": 0.19696056395769118,
      "num_tokens": 15925439.0,
      "step": 6945
    },
    {
      "entropy": 5.289969635009766,
      "epoch": 0.6676272814601345,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004963126248968087,
      "loss": 5.0201,
      "mean_token_accuracy": 0.22049615681171417,
      "num_tokens": 15937762.0,
      "step": 6950
    },
    {
      "entropy": 5.265688419342041,
      "epoch": 0.6681075888568684,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004963064410026577,
      "loss": 5.0547,
      "mean_token_accuracy": 0.21403318494558335,
      "num_tokens": 15948656.0,
      "step": 6955
    },
    {
      "entropy": 5.391811513900757,
      "epoch": 0.6685878962536023,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004963002519704281,
      "loss": 5.1334,
      "mean_token_accuracy": 0.20170396864414214,
      "num_tokens": 15960376.0,
      "step": 6960
    },
    {
      "entropy": 5.384323215484619,
      "epoch": 0.6690682036503363,
      "grad_norm": 0.96875,
      "learning_rate": 0.000496294057800264,
      "loss": 5.145,
      "mean_token_accuracy": 0.20424775034189224,
      "num_tokens": 15971913.0,
      "step": 6965
    },
    {
      "entropy": 5.390318536758423,
      "epoch": 0.6695485110470701,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004962878584923089,
      "loss": 5.2061,
      "mean_token_accuracy": 0.19896737337112427,
      "num_tokens": 15984775.0,
      "step": 6970
    },
    {
      "entropy": 5.364650821685791,
      "epoch": 0.670028818443804,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004962816540467068,
      "loss": 5.1338,
      "mean_token_accuracy": 0.2048588514328003,
      "num_tokens": 15996717.0,
      "step": 6975
    },
    {
      "entropy": 5.3452088832855225,
      "epoch": 0.6705091258405379,
      "grad_norm": 1.0,
      "learning_rate": 0.0004962754444636017,
      "loss": 5.1314,
      "mean_token_accuracy": 0.2110610857605934,
      "num_tokens": 16007964.0,
      "step": 6980
    },
    {
      "entropy": 5.322988605499267,
      "epoch": 0.6709894332372719,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000496269229743138,
      "loss": 5.1443,
      "mean_token_accuracy": 0.20443421453237534,
      "num_tokens": 16019185.0,
      "step": 6985
    },
    {
      "entropy": 5.421507120132446,
      "epoch": 0.6714697406340058,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004962630098854597,
      "loss": 5.1381,
      "mean_token_accuracy": 0.20968109518289565,
      "num_tokens": 16029983.0,
      "step": 6990
    },
    {
      "entropy": 5.3502603530883786,
      "epoch": 0.6719500480307397,
      "grad_norm": 0.875,
      "learning_rate": 0.0004962567848907113,
      "loss": 5.0941,
      "mean_token_accuracy": 0.2066495418548584,
      "num_tokens": 16040574.0,
      "step": 6995
    },
    {
      "entropy": 5.393819761276245,
      "epoch": 0.6724303554274735,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962505547590374,
      "loss": 5.2385,
      "mean_token_accuracy": 0.20304362177848817,
      "num_tokens": 16052037.0,
      "step": 7000
    },
    {
      "entropy": 5.446655702590943,
      "epoch": 0.6729106628242075,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004962443194905826,
      "loss": 5.2233,
      "mean_token_accuracy": 0.2005530834197998,
      "num_tokens": 16063413.0,
      "step": 7005
    },
    {
      "entropy": 5.397491121292115,
      "epoch": 0.6733909702209414,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004962380790854916,
      "loss": 5.1051,
      "mean_token_accuracy": 0.2054908126592636,
      "num_tokens": 16074373.0,
      "step": 7010
    },
    {
      "entropy": 5.419117069244384,
      "epoch": 0.6738712776176753,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004962318335439094,
      "loss": 5.2461,
      "mean_token_accuracy": 0.19945778101682662,
      "num_tokens": 16086575.0,
      "step": 7015
    },
    {
      "entropy": 5.350155591964722,
      "epoch": 0.6743515850144092,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004962255828659809,
      "loss": 5.0898,
      "mean_token_accuracy": 0.20805711746215821,
      "num_tokens": 16098529.0,
      "step": 7020
    },
    {
      "entropy": 5.310775423049927,
      "epoch": 0.6748318924111432,
      "grad_norm": 0.875,
      "learning_rate": 0.0004962193270518513,
      "loss": 5.1478,
      "mean_token_accuracy": 0.20812925547361374,
      "num_tokens": 16110085.0,
      "step": 7025
    },
    {
      "entropy": 5.348368215560913,
      "epoch": 0.675312199807877,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004962130661016659,
      "loss": 5.0326,
      "mean_token_accuracy": 0.20979710072278976,
      "num_tokens": 16120249.0,
      "step": 7030
    },
    {
      "entropy": 5.436573314666748,
      "epoch": 0.6757925072046109,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004962068000155699,
      "loss": 5.3062,
      "mean_token_accuracy": 0.1937258645892143,
      "num_tokens": 16132645.0,
      "step": 7035
    },
    {
      "entropy": 5.3813276290893555,
      "epoch": 0.6762728146013448,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004962005287937088,
      "loss": 5.1883,
      "mean_token_accuracy": 0.20899680703878404,
      "num_tokens": 16143808.0,
      "step": 7040
    },
    {
      "entropy": 5.345149374008178,
      "epoch": 0.6767531219980788,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004961942524362283,
      "loss": 5.1813,
      "mean_token_accuracy": 0.2076207548379898,
      "num_tokens": 16154309.0,
      "step": 7045
    },
    {
      "entropy": 5.362104892730713,
      "epoch": 0.6772334293948127,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004961879709432741,
      "loss": 5.1443,
      "mean_token_accuracy": 0.20428162813186646,
      "num_tokens": 16164654.0,
      "step": 7050
    },
    {
      "entropy": 5.4138587474822994,
      "epoch": 0.6777137367915466,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496181684314992,
      "loss": 5.1889,
      "mean_token_accuracy": 0.20368209928274156,
      "num_tokens": 16177837.0,
      "step": 7055
    },
    {
      "entropy": 5.327811813354492,
      "epoch": 0.6781940441882806,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004961753925515279,
      "loss": 5.071,
      "mean_token_accuracy": 0.21381984055042266,
      "num_tokens": 16189073.0,
      "step": 7060
    },
    {
      "entropy": 5.441262292861938,
      "epoch": 0.6786743515850144,
      "grad_norm": 1.0,
      "learning_rate": 0.000496169095653028,
      "loss": 5.2629,
      "mean_token_accuracy": 0.2001970276236534,
      "num_tokens": 16200371.0,
      "step": 7065
    },
    {
      "entropy": 5.345044231414795,
      "epoch": 0.6791546589817483,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004961627936196384,
      "loss": 5.047,
      "mean_token_accuracy": 0.21157737970352172,
      "num_tokens": 16210526.0,
      "step": 7070
    },
    {
      "entropy": 5.349695062637329,
      "epoch": 0.6796349663784822,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004961564864515055,
      "loss": 5.1569,
      "mean_token_accuracy": 0.20670645087957382,
      "num_tokens": 16221687.0,
      "step": 7075
    },
    {
      "entropy": 5.292357587814331,
      "epoch": 0.6801152737752162,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004961501741487757,
      "loss": 5.0413,
      "mean_token_accuracy": 0.22109192609786987,
      "num_tokens": 16233828.0,
      "step": 7080
    },
    {
      "entropy": 5.325454235076904,
      "epoch": 0.6805955811719501,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004961438567115955,
      "loss": 5.0266,
      "mean_token_accuracy": 0.22332497239112853,
      "num_tokens": 16243900.0,
      "step": 7085
    },
    {
      "entropy": 5.273006868362427,
      "epoch": 0.681075888568684,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004961375341401116,
      "loss": 5.1255,
      "mean_token_accuracy": 0.2100134715437889,
      "num_tokens": 16256347.0,
      "step": 7090
    },
    {
      "entropy": 5.345168590545654,
      "epoch": 0.6815561959654178,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004961312064344708,
      "loss": 5.0347,
      "mean_token_accuracy": 0.2135253593325615,
      "num_tokens": 16267743.0,
      "step": 7095
    },
    {
      "entropy": 5.28406720161438,
      "epoch": 0.6820365033621518,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00049612487359482,
      "loss": 5.0776,
      "mean_token_accuracy": 0.21229360103607178,
      "num_tokens": 16278280.0,
      "step": 7100
    },
    {
      "entropy": 5.274214792251587,
      "epoch": 0.6825168107588857,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004961185356213062,
      "loss": 5.0636,
      "mean_token_accuracy": 0.21724810749292373,
      "num_tokens": 16288568.0,
      "step": 7105
    },
    {
      "entropy": 5.32531156539917,
      "epoch": 0.6829971181556196,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004961121925140767,
      "loss": 5.152,
      "mean_token_accuracy": 0.21107460260391236,
      "num_tokens": 16300730.0,
      "step": 7110
    },
    {
      "entropy": 5.359479808807373,
      "epoch": 0.6834774255523535,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004961058442732786,
      "loss": 5.1632,
      "mean_token_accuracy": 0.20375294834375382,
      "num_tokens": 16311789.0,
      "step": 7115
    },
    {
      "entropy": 5.325752592086792,
      "epoch": 0.6839577329490875,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004960994908990594,
      "loss": 5.1131,
      "mean_token_accuracy": 0.20331291407346724,
      "num_tokens": 16324439.0,
      "step": 7120
    },
    {
      "entropy": 5.438413667678833,
      "epoch": 0.6844380403458213,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004960931323915665,
      "loss": 5.2556,
      "mean_token_accuracy": 0.2140980541706085,
      "num_tokens": 16335344.0,
      "step": 7125
    },
    {
      "entropy": 5.333748769760132,
      "epoch": 0.6849183477425552,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004960867687509475,
      "loss": 5.0934,
      "mean_token_accuracy": 0.21208712458610535,
      "num_tokens": 16349018.0,
      "step": 7130
    },
    {
      "entropy": 5.390694570541382,
      "epoch": 0.6853986551392891,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004960803999773504,
      "loss": 5.1762,
      "mean_token_accuracy": 0.20469519644975662,
      "num_tokens": 16360137.0,
      "step": 7135
    },
    {
      "entropy": 5.427073669433594,
      "epoch": 0.6858789625360231,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004960740260709228,
      "loss": 5.182,
      "mean_token_accuracy": 0.2063075363636017,
      "num_tokens": 16372277.0,
      "step": 7140
    },
    {
      "entropy": 5.40649676322937,
      "epoch": 0.686359269932757,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004960676470318128,
      "loss": 5.1838,
      "mean_token_accuracy": 0.20653775185346604,
      "num_tokens": 16383440.0,
      "step": 7145
    },
    {
      "entropy": 5.370810794830322,
      "epoch": 0.6868395773294909,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004960612628601683,
      "loss": 5.1479,
      "mean_token_accuracy": 0.20952605754137038,
      "num_tokens": 16394330.0,
      "step": 7150
    },
    {
      "entropy": 5.262090873718262,
      "epoch": 0.6873198847262247,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004960548735561379,
      "loss": 5.0803,
      "mean_token_accuracy": 0.2153622731566429,
      "num_tokens": 16405734.0,
      "step": 7155
    },
    {
      "entropy": 5.348772668838501,
      "epoch": 0.6878001921229587,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004960484791198697,
      "loss": 5.1016,
      "mean_token_accuracy": 0.2042395666241646,
      "num_tokens": 16416025.0,
      "step": 7160
    },
    {
      "entropy": 5.354042959213257,
      "epoch": 0.6882804995196926,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004960420795515121,
      "loss": 5.1131,
      "mean_token_accuracy": 0.20813387930393218,
      "num_tokens": 16427416.0,
      "step": 7165
    },
    {
      "entropy": 5.32905535697937,
      "epoch": 0.6887608069164265,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004960356748512138,
      "loss": 5.1023,
      "mean_token_accuracy": 0.21077259182929992,
      "num_tokens": 16438073.0,
      "step": 7170
    },
    {
      "entropy": 5.315211772918701,
      "epoch": 0.6892411143131604,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004960292650191236,
      "loss": 5.0462,
      "mean_token_accuracy": 0.20846395790576935,
      "num_tokens": 16449994.0,
      "step": 7175
    },
    {
      "entropy": 5.308739948272705,
      "epoch": 0.6897214217098944,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004960228500553899,
      "loss": 5.0613,
      "mean_token_accuracy": 0.22385766804218293,
      "num_tokens": 16460355.0,
      "step": 7180
    },
    {
      "entropy": 5.411591672897339,
      "epoch": 0.6902017291066282,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004960164299601623,
      "loss": 5.1416,
      "mean_token_accuracy": 0.20120686888694764,
      "num_tokens": 16472274.0,
      "step": 7185
    },
    {
      "entropy": 5.42814679145813,
      "epoch": 0.6906820365033621,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004960100047335892,
      "loss": 5.1539,
      "mean_token_accuracy": 0.2094866544008255,
      "num_tokens": 16482959.0,
      "step": 7190
    },
    {
      "entropy": 5.357362604141235,
      "epoch": 0.691162343900096,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004960035743758202,
      "loss": 5.1985,
      "mean_token_accuracy": 0.2041933611035347,
      "num_tokens": 16494533.0,
      "step": 7195
    },
    {
      "entropy": 5.262730598449707,
      "epoch": 0.69164265129683,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004959971388870044,
      "loss": 5.0605,
      "mean_token_accuracy": 0.2141747683286667,
      "num_tokens": 16506124.0,
      "step": 7200
    },
    {
      "entropy": 5.33617844581604,
      "epoch": 0.6921229586935639,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004959906982672912,
      "loss": 5.0827,
      "mean_token_accuracy": 0.20826901495456696,
      "num_tokens": 16517867.0,
      "step": 7205
    },
    {
      "entropy": 5.284541940689087,
      "epoch": 0.6926032660902978,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004959842525168302,
      "loss": 5.0266,
      "mean_token_accuracy": 0.2125307008624077,
      "num_tokens": 16529075.0,
      "step": 7210
    },
    {
      "entropy": 5.354151201248169,
      "epoch": 0.6930835734870316,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004959778016357712,
      "loss": 5.0662,
      "mean_token_accuracy": 0.21161008030176162,
      "num_tokens": 16540326.0,
      "step": 7215
    },
    {
      "entropy": 5.327894973754883,
      "epoch": 0.6935638808837656,
      "grad_norm": 0.875,
      "learning_rate": 0.0004959713456242637,
      "loss": 5.1152,
      "mean_token_accuracy": 0.20628926604986192,
      "num_tokens": 16551570.0,
      "step": 7220
    },
    {
      "entropy": 5.449658012390136,
      "epoch": 0.6940441882804995,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004959648844824576,
      "loss": 5.1542,
      "mean_token_accuracy": 0.2022443100810051,
      "num_tokens": 16562636.0,
      "step": 7225
    },
    {
      "entropy": 5.3453999042510985,
      "epoch": 0.6945244956772334,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004959584182105032,
      "loss": 5.0464,
      "mean_token_accuracy": 0.21254502683877946,
      "num_tokens": 16573867.0,
      "step": 7230
    },
    {
      "entropy": 5.264127635955811,
      "epoch": 0.6950048030739674,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004959519468085504,
      "loss": 5.0547,
      "mean_token_accuracy": 0.21692499071359633,
      "num_tokens": 16584533.0,
      "step": 7235
    },
    {
      "entropy": 5.315955066680909,
      "epoch": 0.6954851104707013,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004959454702767494,
      "loss": 5.2384,
      "mean_token_accuracy": 0.21240682601928712,
      "num_tokens": 16596562.0,
      "step": 7240
    },
    {
      "entropy": 5.3698793888092045,
      "epoch": 0.6959654178674352,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004959389886152507,
      "loss": 5.0978,
      "mean_token_accuracy": 0.215421724319458,
      "num_tokens": 16607508.0,
      "step": 7245
    },
    {
      "entropy": 5.389533185958863,
      "epoch": 0.696445725264169,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004959325018242048,
      "loss": 5.1712,
      "mean_token_accuracy": 0.20800974667072297,
      "num_tokens": 16617737.0,
      "step": 7250
    },
    {
      "entropy": 5.383937072753906,
      "epoch": 0.696926032660903,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004959260099037622,
      "loss": 5.1756,
      "mean_token_accuracy": 0.206690476834774,
      "num_tokens": 16628518.0,
      "step": 7255
    },
    {
      "entropy": 5.391141510009765,
      "epoch": 0.6974063400576369,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004959195128540737,
      "loss": 5.1937,
      "mean_token_accuracy": 0.20455787181854249,
      "num_tokens": 16639644.0,
      "step": 7260
    },
    {
      "entropy": 5.479088926315308,
      "epoch": 0.6978866474543708,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049591301067529,
      "loss": 5.2483,
      "mean_token_accuracy": 0.19821529239416122,
      "num_tokens": 16650882.0,
      "step": 7265
    },
    {
      "entropy": 5.385912227630615,
      "epoch": 0.6983669548511047,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004959065033675623,
      "loss": 5.133,
      "mean_token_accuracy": 0.20692792683839797,
      "num_tokens": 16662464.0,
      "step": 7270
    },
    {
      "entropy": 5.374574375152588,
      "epoch": 0.6988472622478387,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004958999909310414,
      "loss": 5.1783,
      "mean_token_accuracy": 0.20143779665231704,
      "num_tokens": 16673718.0,
      "step": 7275
    },
    {
      "entropy": 5.3214654445648195,
      "epoch": 0.6993275696445725,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004958934733658788,
      "loss": 5.1055,
      "mean_token_accuracy": 0.20703895539045333,
      "num_tokens": 16684957.0,
      "step": 7280
    },
    {
      "entropy": 5.380169534683228,
      "epoch": 0.6998078770413064,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004958869506722256,
      "loss": 5.1488,
      "mean_token_accuracy": 0.20702529996633529,
      "num_tokens": 16695782.0,
      "step": 7285
    },
    {
      "entropy": 5.327401781082154,
      "epoch": 0.7002881844380403,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004958804228502332,
      "loss": 5.0595,
      "mean_token_accuracy": 0.21661947667598724,
      "num_tokens": 16707448.0,
      "step": 7290
    },
    {
      "entropy": 5.26526985168457,
      "epoch": 0.7007684918347743,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004958738899000534,
      "loss": 5.0817,
      "mean_token_accuracy": 0.20653405636548997,
      "num_tokens": 16718074.0,
      "step": 7295
    },
    {
      "entropy": 5.367698526382446,
      "epoch": 0.7012487992315082,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004958673518218377,
      "loss": 5.1174,
      "mean_token_accuracy": 0.21311235874891282,
      "num_tokens": 16728656.0,
      "step": 7300
    },
    {
      "entropy": 5.3638927936553955,
      "epoch": 0.7017291066282421,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004958608086157379,
      "loss": 5.179,
      "mean_token_accuracy": 0.20986603200435638,
      "num_tokens": 16740703.0,
      "step": 7305
    },
    {
      "entropy": 5.32499327659607,
      "epoch": 0.7022094140249759,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000495854260281906,
      "loss": 5.0442,
      "mean_token_accuracy": 0.22114641666412355,
      "num_tokens": 16753320.0,
      "step": 7310
    },
    {
      "entropy": 5.272854995727539,
      "epoch": 0.7026897214217099,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004958477068204941,
      "loss": 5.1177,
      "mean_token_accuracy": 0.21227967739105225,
      "num_tokens": 16764889.0,
      "step": 7315
    },
    {
      "entropy": 5.435892820358276,
      "epoch": 0.7031700288184438,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000495841148231654,
      "loss": 5.1535,
      "mean_token_accuracy": 0.2119811773300171,
      "num_tokens": 16776301.0,
      "step": 7320
    },
    {
      "entropy": 5.379011058807373,
      "epoch": 0.7036503362151777,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004958345845155383,
      "loss": 5.1393,
      "mean_token_accuracy": 0.2159304976463318,
      "num_tokens": 16786935.0,
      "step": 7325
    },
    {
      "entropy": 5.304282379150391,
      "epoch": 0.7041306436119116,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004958280156722992,
      "loss": 5.1395,
      "mean_token_accuracy": 0.20692165046930314,
      "num_tokens": 16799335.0,
      "step": 7330
    },
    {
      "entropy": 5.264208889007568,
      "epoch": 0.7046109510086456,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004958214417020894,
      "loss": 5.0824,
      "mean_token_accuracy": 0.20889249294996262,
      "num_tokens": 16811376.0,
      "step": 7335
    },
    {
      "entropy": 5.422280263900757,
      "epoch": 0.7050912584053795,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004958148626050614,
      "loss": 5.1378,
      "mean_token_accuracy": 0.20535912811756135,
      "num_tokens": 16822859.0,
      "step": 7340
    },
    {
      "entropy": 5.3414459228515625,
      "epoch": 0.7055715658021133,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000495808278381368,
      "loss": 5.1228,
      "mean_token_accuracy": 0.20882423967123032,
      "num_tokens": 16833831.0,
      "step": 7345
    },
    {
      "entropy": 5.337645864486694,
      "epoch": 0.7060518731988472,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000495801689031162,
      "loss": 5.1459,
      "mean_token_accuracy": 0.20560007393360138,
      "num_tokens": 16846827.0,
      "step": 7350
    },
    {
      "entropy": 5.391799640655518,
      "epoch": 0.7065321805955812,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004957950945545965,
      "loss": 5.1186,
      "mean_token_accuracy": 0.20683661997318267,
      "num_tokens": 16858166.0,
      "step": 7355
    },
    {
      "entropy": 5.412299585342407,
      "epoch": 0.7070124879923151,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004957884949518246,
      "loss": 5.2476,
      "mean_token_accuracy": 0.19617687314748763,
      "num_tokens": 16870201.0,
      "step": 7360
    },
    {
      "entropy": 5.411854648590088,
      "epoch": 0.707492795389049,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957818902229992,
      "loss": 5.1211,
      "mean_token_accuracy": 0.2096790924668312,
      "num_tokens": 16880891.0,
      "step": 7365
    },
    {
      "entropy": 5.396496438980103,
      "epoch": 0.7079731027857828,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004957752803682741,
      "loss": 5.1882,
      "mean_token_accuracy": 0.20081070214509963,
      "num_tokens": 16893498.0,
      "step": 7370
    },
    {
      "entropy": 5.39955735206604,
      "epoch": 0.7084534101825168,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004957686653878024,
      "loss": 5.1508,
      "mean_token_accuracy": 0.2093782216310501,
      "num_tokens": 16904959.0,
      "step": 7375
    },
    {
      "entropy": 5.398483228683472,
      "epoch": 0.7089337175792507,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495762045281738,
      "loss": 5.1832,
      "mean_token_accuracy": 0.2067160815000534,
      "num_tokens": 16916547.0,
      "step": 7380
    },
    {
      "entropy": 5.311323881149292,
      "epoch": 0.7094140249759846,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004957554200502344,
      "loss": 5.1255,
      "mean_token_accuracy": 0.21099474132061005,
      "num_tokens": 16928580.0,
      "step": 7385
    },
    {
      "entropy": 5.380280160903931,
      "epoch": 0.7098943323727186,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004957487896934454,
      "loss": 5.1512,
      "mean_token_accuracy": 0.20483650714159013,
      "num_tokens": 16941247.0,
      "step": 7390
    },
    {
      "entropy": 5.442416191101074,
      "epoch": 0.7103746397694525,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000495742154211525,
      "loss": 5.1757,
      "mean_token_accuracy": 0.20876175463199614,
      "num_tokens": 16951119.0,
      "step": 7395
    },
    {
      "entropy": 5.311107206344604,
      "epoch": 0.7108549471661864,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004957355136046272,
      "loss": 5.0984,
      "mean_token_accuracy": 0.20989696085453033,
      "num_tokens": 16962608.0,
      "step": 7400
    },
    {
      "entropy": 5.298178100585938,
      "epoch": 0.7113352545629202,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004957288678729064,
      "loss": 5.1248,
      "mean_token_accuracy": 0.2083708956837654,
      "num_tokens": 16973291.0,
      "step": 7405
    },
    {
      "entropy": 5.409120750427246,
      "epoch": 0.7118155619596542,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004957222170165166,
      "loss": 5.153,
      "mean_token_accuracy": 0.20308410227298737,
      "num_tokens": 16984895.0,
      "step": 7410
    },
    {
      "entropy": 5.399225854873658,
      "epoch": 0.7122958693563881,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004957155610356124,
      "loss": 5.1601,
      "mean_token_accuracy": 0.20235242545604706,
      "num_tokens": 16997109.0,
      "step": 7415
    },
    {
      "entropy": 5.323979473114013,
      "epoch": 0.712776176753122,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004957088999303481,
      "loss": 5.1081,
      "mean_token_accuracy": 0.2076752156019211,
      "num_tokens": 17008640.0,
      "step": 7420
    },
    {
      "entropy": 5.348388385772705,
      "epoch": 0.7132564841498559,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004957022337008787,
      "loss": 5.0782,
      "mean_token_accuracy": 0.21174404621124268,
      "num_tokens": 17018924.0,
      "step": 7425
    },
    {
      "entropy": 5.319825458526611,
      "epoch": 0.7137367915465899,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004956955623473587,
      "loss": 5.1171,
      "mean_token_accuracy": 0.20976233035326003,
      "num_tokens": 17029932.0,
      "step": 7430
    },
    {
      "entropy": 5.389454078674317,
      "epoch": 0.7142170989433237,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000495688885869943,
      "loss": 5.2015,
      "mean_token_accuracy": 0.20615289062261583,
      "num_tokens": 17042164.0,
      "step": 7435
    },
    {
      "entropy": 5.367098999023438,
      "epoch": 0.7146974063400576,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004956822042687868,
      "loss": 5.174,
      "mean_token_accuracy": 0.20286040604114533,
      "num_tokens": 17054729.0,
      "step": 7440
    },
    {
      "entropy": 5.296559524536133,
      "epoch": 0.7151777137367915,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004956755175440451,
      "loss": 5.0446,
      "mean_token_accuracy": 0.22101425379514694,
      "num_tokens": 17066537.0,
      "step": 7445
    },
    {
      "entropy": 5.37962327003479,
      "epoch": 0.7156580211335255,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004956688256958731,
      "loss": 5.0988,
      "mean_token_accuracy": 0.21140699386596679,
      "num_tokens": 17077787.0,
      "step": 7450
    },
    {
      "entropy": 5.319226169586182,
      "epoch": 0.7161383285302594,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004956621287244262,
      "loss": 5.1359,
      "mean_token_accuracy": 0.20658225268125535,
      "num_tokens": 17089555.0,
      "step": 7455
    },
    {
      "entropy": 5.406843376159668,
      "epoch": 0.7166186359269933,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004956554266298599,
      "loss": 5.093,
      "mean_token_accuracy": 0.2141314923763275,
      "num_tokens": 17099911.0,
      "step": 7460
    },
    {
      "entropy": 5.288338327407837,
      "epoch": 0.7170989433237271,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004956487194123298,
      "loss": 5.0853,
      "mean_token_accuracy": 0.20724095404148102,
      "num_tokens": 17110880.0,
      "step": 7465
    },
    {
      "entropy": 5.336555480957031,
      "epoch": 0.7175792507204611,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004956420070719918,
      "loss": 5.0505,
      "mean_token_accuracy": 0.2131749987602234,
      "num_tokens": 17122272.0,
      "step": 7470
    },
    {
      "entropy": 5.3773088455200195,
      "epoch": 0.718059558117195,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004956352896090014,
      "loss": 5.1761,
      "mean_token_accuracy": 0.2062709778547287,
      "num_tokens": 17133995.0,
      "step": 7475
    },
    {
      "entropy": 5.381700038909912,
      "epoch": 0.7185398655139289,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004956285670235147,
      "loss": 5.0493,
      "mean_token_accuracy": 0.2148824781179428,
      "num_tokens": 17145970.0,
      "step": 7480
    },
    {
      "entropy": 5.41835470199585,
      "epoch": 0.7190201729106628,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004956218393156879,
      "loss": 5.2548,
      "mean_token_accuracy": 0.20228676795959472,
      "num_tokens": 17157747.0,
      "step": 7485
    },
    {
      "entropy": 5.330062198638916,
      "epoch": 0.7195004803073968,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004956151064856772,
      "loss": 5.188,
      "mean_token_accuracy": 0.20652631521224976,
      "num_tokens": 17168357.0,
      "step": 7490
    },
    {
      "entropy": 5.374532318115234,
      "epoch": 0.7199807877041307,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004956083685336386,
      "loss": 5.1299,
      "mean_token_accuracy": 0.21559945046901702,
      "num_tokens": 17179871.0,
      "step": 7495
    },
    {
      "entropy": 5.402276611328125,
      "epoch": 0.7204610951008645,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004956016254597289,
      "loss": 5.0839,
      "mean_token_accuracy": 0.21271447241306304,
      "num_tokens": 17190456.0,
      "step": 7500
    },
    {
      "entropy": 5.321882534027099,
      "epoch": 0.7209414024975984,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955948772641044,
      "loss": 5.1747,
      "mean_token_accuracy": 0.2006893962621689,
      "num_tokens": 17201623.0,
      "step": 7505
    },
    {
      "entropy": 5.309241771697998,
      "epoch": 0.7214217098943324,
      "grad_norm": 0.953125,
      "learning_rate": 0.000495588123946922,
      "loss": 5.0703,
      "mean_token_accuracy": 0.21214835047721864,
      "num_tokens": 17212891.0,
      "step": 7510
    },
    {
      "entropy": 5.343848085403442,
      "epoch": 0.7219020172910663,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955813655083384,
      "loss": 5.1317,
      "mean_token_accuracy": 0.2064789801836014,
      "num_tokens": 17223983.0,
      "step": 7515
    },
    {
      "entropy": 5.363732385635376,
      "epoch": 0.7223823246878002,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955746019485104,
      "loss": 5.1656,
      "mean_token_accuracy": 0.20909778773784637,
      "num_tokens": 17233701.0,
      "step": 7520
    },
    {
      "entropy": 5.352098035812378,
      "epoch": 0.722862632084534,
      "grad_norm": 0.875,
      "learning_rate": 0.000495567833267595,
      "loss": 5.1725,
      "mean_token_accuracy": 0.20591531246900557,
      "num_tokens": 17245958.0,
      "step": 7525
    },
    {
      "entropy": 5.470615196228027,
      "epoch": 0.723342939481268,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004955610594657496,
      "loss": 5.1483,
      "mean_token_accuracy": 0.20365383177995683,
      "num_tokens": 17257687.0,
      "step": 7530
    },
    {
      "entropy": 5.394668102264404,
      "epoch": 0.7238232468780019,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004955542805431313,
      "loss": 5.2072,
      "mean_token_accuracy": 0.20809406340122222,
      "num_tokens": 17270247.0,
      "step": 7535
    },
    {
      "entropy": 5.337820434570313,
      "epoch": 0.7243035542747358,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004955474964998976,
      "loss": 5.1177,
      "mean_token_accuracy": 0.21522586792707443,
      "num_tokens": 17281920.0,
      "step": 7540
    },
    {
      "entropy": 5.406730318069458,
      "epoch": 0.7247838616714697,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004955407073362058,
      "loss": 5.281,
      "mean_token_accuracy": 0.19879963546991347,
      "num_tokens": 17293602.0,
      "step": 7545
    },
    {
      "entropy": 5.429531192779541,
      "epoch": 0.7252641690682037,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004955339130522136,
      "loss": 5.0934,
      "mean_token_accuracy": 0.21053230464458467,
      "num_tokens": 17304484.0,
      "step": 7550
    },
    {
      "entropy": 5.271792984008789,
      "epoch": 0.7257444764649376,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000495527113648079,
      "loss": 5.028,
      "mean_token_accuracy": 0.2155469089746475,
      "num_tokens": 17314953.0,
      "step": 7555
    },
    {
      "entropy": 5.345812654495239,
      "epoch": 0.7262247838616714,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004955203091239596,
      "loss": 5.146,
      "mean_token_accuracy": 0.20928613245487213,
      "num_tokens": 17326813.0,
      "step": 7560
    },
    {
      "entropy": 5.449349451065063,
      "epoch": 0.7267050912584054,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004955134994800134,
      "loss": 5.2251,
      "mean_token_accuracy": 0.20042096823453903,
      "num_tokens": 17337968.0,
      "step": 7565
    },
    {
      "entropy": 5.375257968902588,
      "epoch": 0.7271853986551393,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004955066847163984,
      "loss": 5.1194,
      "mean_token_accuracy": 0.20437183529138564,
      "num_tokens": 17350406.0,
      "step": 7570
    },
    {
      "entropy": 5.28691668510437,
      "epoch": 0.7276657060518732,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004954998648332731,
      "loss": 5.0394,
      "mean_token_accuracy": 0.213681223988533,
      "num_tokens": 17361888.0,
      "step": 7575
    },
    {
      "entropy": 5.3225846767425535,
      "epoch": 0.7281460134486071,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004954930398307956,
      "loss": 5.1517,
      "mean_token_accuracy": 0.2064734071493149,
      "num_tokens": 17374047.0,
      "step": 7580
    },
    {
      "entropy": 5.391057634353638,
      "epoch": 0.7286263208453411,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004954862097091245,
      "loss": 5.1435,
      "mean_token_accuracy": 0.20068034529685974,
      "num_tokens": 17386175.0,
      "step": 7585
    },
    {
      "entropy": 5.3398766040802,
      "epoch": 0.729106628242075,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004954793744684184,
      "loss": 5.1413,
      "mean_token_accuracy": 0.21976462602615357,
      "num_tokens": 17398168.0,
      "step": 7590
    },
    {
      "entropy": 5.304914951324463,
      "epoch": 0.7295869356388088,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004954725341088358,
      "loss": 5.0376,
      "mean_token_accuracy": 0.2142294764518738,
      "num_tokens": 17408825.0,
      "step": 7595
    },
    {
      "entropy": 5.323363447189331,
      "epoch": 0.7300672430355427,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004954656886305356,
      "loss": 5.1357,
      "mean_token_accuracy": 0.2089396059513092,
      "num_tokens": 17419813.0,
      "step": 7600
    },
    {
      "entropy": 5.409443712234497,
      "epoch": 0.7305475504322767,
      "grad_norm": 1.0,
      "learning_rate": 0.0004954588380336768,
      "loss": 5.1802,
      "mean_token_accuracy": 0.20178788006305695,
      "num_tokens": 17431134.0,
      "step": 7605
    },
    {
      "entropy": 5.310650157928467,
      "epoch": 0.7310278578290106,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004954519823184184,
      "loss": 5.0779,
      "mean_token_accuracy": 0.21509800106287003,
      "num_tokens": 17442614.0,
      "step": 7610
    },
    {
      "entropy": 5.380193662643433,
      "epoch": 0.7315081652257445,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004954451214849196,
      "loss": 5.0713,
      "mean_token_accuracy": 0.2136652797460556,
      "num_tokens": 17454615.0,
      "step": 7615
    },
    {
      "entropy": 5.363864898681641,
      "epoch": 0.7319884726224783,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004954382555333397,
      "loss": 5.1193,
      "mean_token_accuracy": 0.20585674196481704,
      "num_tokens": 17467025.0,
      "step": 7620
    },
    {
      "entropy": 5.318731307983398,
      "epoch": 0.7324687800192123,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000495431384463838,
      "loss": 5.1011,
      "mean_token_accuracy": 0.21396268010139466,
      "num_tokens": 17477681.0,
      "step": 7625
    },
    {
      "entropy": 5.366606712341309,
      "epoch": 0.7329490874159462,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004954245082765741,
      "loss": 5.1476,
      "mean_token_accuracy": 0.2102656289935112,
      "num_tokens": 17489814.0,
      "step": 7630
    },
    {
      "entropy": 5.36464204788208,
      "epoch": 0.7334293948126801,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004954176269717077,
      "loss": 5.1303,
      "mean_token_accuracy": 0.21068463772535323,
      "num_tokens": 17501701.0,
      "step": 7635
    },
    {
      "entropy": 5.336238145828247,
      "epoch": 0.733909702209414,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004954107405493984,
      "loss": 5.0546,
      "mean_token_accuracy": 0.21546552926301957,
      "num_tokens": 17513585.0,
      "step": 7640
    },
    {
      "entropy": 5.389128828048706,
      "epoch": 0.734390009606148,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004954038490098064,
      "loss": 5.1881,
      "mean_token_accuracy": 0.21105958968400956,
      "num_tokens": 17525557.0,
      "step": 7645
    },
    {
      "entropy": 5.282487583160401,
      "epoch": 0.7348703170028819,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004953969523530914,
      "loss": 5.0427,
      "mean_token_accuracy": 0.21114206463098525,
      "num_tokens": 17538312.0,
      "step": 7650
    },
    {
      "entropy": 5.4409853458404545,
      "epoch": 0.7353506243996157,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004953900505794136,
      "loss": 5.1893,
      "mean_token_accuracy": 0.21002898812294007,
      "num_tokens": 17550248.0,
      "step": 7655
    },
    {
      "entropy": 5.363926839828491,
      "epoch": 0.7358309317963496,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004953831436889332,
      "loss": 5.0502,
      "mean_token_accuracy": 0.20910338908433915,
      "num_tokens": 17560624.0,
      "step": 7660
    },
    {
      "entropy": 5.35666708946228,
      "epoch": 0.7363112391930836,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004953762316818106,
      "loss": 5.0973,
      "mean_token_accuracy": 0.20386057198047638,
      "num_tokens": 17572439.0,
      "step": 7665
    },
    {
      "entropy": 5.39898715019226,
      "epoch": 0.7367915465898175,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004953693145582064,
      "loss": 5.1628,
      "mean_token_accuracy": 0.19806077778339387,
      "num_tokens": 17583120.0,
      "step": 7670
    },
    {
      "entropy": 5.249486303329467,
      "epoch": 0.7372718539865514,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000495362392318281,
      "loss": 4.981,
      "mean_token_accuracy": 0.21629206240177154,
      "num_tokens": 17593677.0,
      "step": 7675
    },
    {
      "entropy": 5.32330961227417,
      "epoch": 0.7377521613832853,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004953554649621951,
      "loss": 5.1327,
      "mean_token_accuracy": 0.20416050553321838,
      "num_tokens": 17605180.0,
      "step": 7680
    },
    {
      "entropy": 5.385280323028565,
      "epoch": 0.7382324687800192,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004953485324901098,
      "loss": 5.0911,
      "mean_token_accuracy": 0.21398335546255112,
      "num_tokens": 17617459.0,
      "step": 7685
    },
    {
      "entropy": 5.334705448150634,
      "epoch": 0.7387127761767531,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004953415949021857,
      "loss": 5.0893,
      "mean_token_accuracy": 0.21559322476387024,
      "num_tokens": 17628024.0,
      "step": 7690
    },
    {
      "entropy": 5.294398498535156,
      "epoch": 0.739193083573487,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004953346521985843,
      "loss": 5.1065,
      "mean_token_accuracy": 0.21375662684440613,
      "num_tokens": 17639833.0,
      "step": 7695
    },
    {
      "entropy": 5.39537501335144,
      "epoch": 0.7396733909702209,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004953277043794663,
      "loss": 5.2034,
      "mean_token_accuracy": 0.19957452565431594,
      "num_tokens": 17651057.0,
      "step": 7700
    },
    {
      "entropy": 5.384598684310913,
      "epoch": 0.7401536983669549,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004953207514449933,
      "loss": 5.0888,
      "mean_token_accuracy": 0.21481214612722396,
      "num_tokens": 17662288.0,
      "step": 7705
    },
    {
      "entropy": 5.316509485244751,
      "epoch": 0.7406340057636888,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004953137933953267,
      "loss": 5.0918,
      "mean_token_accuracy": 0.20933466255664826,
      "num_tokens": 17673885.0,
      "step": 7710
    },
    {
      "entropy": 5.32520809173584,
      "epoch": 0.7411143131604226,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000495306830230628,
      "loss": 5.1392,
      "mean_token_accuracy": 0.20836299061775207,
      "num_tokens": 17685030.0,
      "step": 7715
    },
    {
      "entropy": 5.323816967010498,
      "epoch": 0.7415946205571565,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004952998619510589,
      "loss": 5.0694,
      "mean_token_accuracy": 0.21306501030921937,
      "num_tokens": 17696624.0,
      "step": 7720
    },
    {
      "entropy": 5.3088788986206055,
      "epoch": 0.7420749279538905,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004952928885567811,
      "loss": 5.0607,
      "mean_token_accuracy": 0.21251924932003022,
      "num_tokens": 17707386.0,
      "step": 7725
    },
    {
      "entropy": 5.451479434967041,
      "epoch": 0.7425552353506244,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004952859100479566,
      "loss": 5.2995,
      "mean_token_accuracy": 0.2028825879096985,
      "num_tokens": 17718605.0,
      "step": 7730
    },
    {
      "entropy": 5.337363195419312,
      "epoch": 0.7430355427473583,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004952789264247474,
      "loss": 5.0926,
      "mean_token_accuracy": 0.20936039686203003,
      "num_tokens": 17730275.0,
      "step": 7735
    },
    {
      "entropy": 5.417365074157715,
      "epoch": 0.7435158501440923,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004952719376873156,
      "loss": 5.1986,
      "mean_token_accuracy": 0.20436775982379912,
      "num_tokens": 17741390.0,
      "step": 7740
    },
    {
      "entropy": 5.347161817550659,
      "epoch": 0.7439961575408262,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004952649438358234,
      "loss": 5.0971,
      "mean_token_accuracy": 0.21377707421779632,
      "num_tokens": 17752354.0,
      "step": 7745
    },
    {
      "entropy": 5.360002899169922,
      "epoch": 0.74447646493756,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004952579448704334,
      "loss": 5.0826,
      "mean_token_accuracy": 0.21161238849163055,
      "num_tokens": 17762839.0,
      "step": 7750
    },
    {
      "entropy": 5.343127012252808,
      "epoch": 0.7449567723342939,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000495250940791308,
      "loss": 5.1677,
      "mean_token_accuracy": 0.20451261699199677,
      "num_tokens": 17775800.0,
      "step": 7755
    },
    {
      "entropy": 5.317103147506714,
      "epoch": 0.7454370797310279,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004952439315986096,
      "loss": 5.0828,
      "mean_token_accuracy": 0.21253036558628083,
      "num_tokens": 17787804.0,
      "step": 7760
    },
    {
      "entropy": 5.436050033569336,
      "epoch": 0.7459173871277618,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952369172925012,
      "loss": 5.1971,
      "mean_token_accuracy": 0.21323916614055632,
      "num_tokens": 17800291.0,
      "step": 7765
    },
    {
      "entropy": 5.342898035049439,
      "epoch": 0.7463976945244957,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004952298978731454,
      "loss": 5.0621,
      "mean_token_accuracy": 0.21446576863527297,
      "num_tokens": 17811548.0,
      "step": 7770
    },
    {
      "entropy": 5.3256432056427006,
      "epoch": 0.7468780019212296,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004952228733407055,
      "loss": 5.0068,
      "mean_token_accuracy": 0.21698374897241593,
      "num_tokens": 17822589.0,
      "step": 7775
    },
    {
      "entropy": 5.278273820877075,
      "epoch": 0.7473583093179635,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004952158436953444,
      "loss": 5.033,
      "mean_token_accuracy": 0.2123190641403198,
      "num_tokens": 17834203.0,
      "step": 7780
    },
    {
      "entropy": 5.319729995727539,
      "epoch": 0.7478386167146974,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004952088089372252,
      "loss": 5.1246,
      "mean_token_accuracy": 0.2076376870274544,
      "num_tokens": 17846238.0,
      "step": 7785
    },
    {
      "entropy": 5.354554319381714,
      "epoch": 0.7483189241114313,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004952017690665114,
      "loss": 5.0316,
      "mean_token_accuracy": 0.21553295552730561,
      "num_tokens": 17857640.0,
      "step": 7790
    },
    {
      "entropy": 5.289981889724731,
      "epoch": 0.7487992315081652,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004951947240833664,
      "loss": 5.0413,
      "mean_token_accuracy": 0.21222982853651046,
      "num_tokens": 17868755.0,
      "step": 7795
    },
    {
      "entropy": 5.303191757202148,
      "epoch": 0.7492795389048992,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004951876739879537,
      "loss": 5.1159,
      "mean_token_accuracy": 0.2043954998254776,
      "num_tokens": 17881078.0,
      "step": 7800
    },
    {
      "entropy": 5.294769525527954,
      "epoch": 0.7497598463016331,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004951806187804371,
      "loss": 5.0614,
      "mean_token_accuracy": 0.21299685537815094,
      "num_tokens": 17893888.0,
      "step": 7805
    },
    {
      "entropy": 5.356993007659912,
      "epoch": 0.7502401536983669,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004951735584609804,
      "loss": 5.1387,
      "mean_token_accuracy": 0.20920005589723586,
      "num_tokens": 17904443.0,
      "step": 7810
    },
    {
      "entropy": 5.303229665756225,
      "epoch": 0.7507204610951008,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004951664930297474,
      "loss": 5.1183,
      "mean_token_accuracy": 0.21148608177900313,
      "num_tokens": 17918090.0,
      "step": 7815
    },
    {
      "entropy": 5.2648204326629635,
      "epoch": 0.7512007684918348,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000495159422486902,
      "loss": 5.067,
      "mean_token_accuracy": 0.2168740600347519,
      "num_tokens": 17929233.0,
      "step": 7820
    },
    {
      "entropy": 5.31578574180603,
      "epoch": 0.7516810758885687,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004951523468326088,
      "loss": 5.101,
      "mean_token_accuracy": 0.20820388942956924,
      "num_tokens": 17940580.0,
      "step": 7825
    },
    {
      "entropy": 5.384409046173095,
      "epoch": 0.7521613832853026,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004951452660670317,
      "loss": 5.1775,
      "mean_token_accuracy": 0.20411736220121385,
      "num_tokens": 17953993.0,
      "step": 7830
    },
    {
      "entropy": 5.3266290664672855,
      "epoch": 0.7526416906820365,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004951381801903352,
      "loss": 5.0279,
      "mean_token_accuracy": 0.20853717923164367,
      "num_tokens": 17966033.0,
      "step": 7835
    },
    {
      "entropy": 5.226248645782471,
      "epoch": 0.7531219980787704,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004951310892026839,
      "loss": 5.0085,
      "mean_token_accuracy": 0.2225278452038765,
      "num_tokens": 17977943.0,
      "step": 7840
    },
    {
      "entropy": 5.333274078369141,
      "epoch": 0.7536023054755043,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004951239931042424,
      "loss": 5.0632,
      "mean_token_accuracy": 0.21436687111854552,
      "num_tokens": 17990135.0,
      "step": 7845
    },
    {
      "entropy": 5.269831848144531,
      "epoch": 0.7540826128722382,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004951168918951753,
      "loss": 5.1028,
      "mean_token_accuracy": 0.21849538534879684,
      "num_tokens": 18002126.0,
      "step": 7850
    },
    {
      "entropy": 5.233481502532959,
      "epoch": 0.7545629202689721,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004951097855756476,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21375206261873245,
      "num_tokens": 18013147.0,
      "step": 7855
    },
    {
      "entropy": 5.379358339309692,
      "epoch": 0.7550432276657061,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004951026741458243,
      "loss": 5.1082,
      "mean_token_accuracy": 0.21287565529346467,
      "num_tokens": 18025146.0,
      "step": 7860
    },
    {
      "entropy": 5.303908967971802,
      "epoch": 0.75552353506244,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004950955576058705,
      "loss": 5.0572,
      "mean_token_accuracy": 0.20765577405691146,
      "num_tokens": 18036412.0,
      "step": 7865
    },
    {
      "entropy": 5.317090034484863,
      "epoch": 0.7560038424591738,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004950884359559513,
      "loss": 5.1284,
      "mean_token_accuracy": 0.21301680952310562,
      "num_tokens": 18048041.0,
      "step": 7870
    },
    {
      "entropy": 5.39552583694458,
      "epoch": 0.7564841498559077,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004950813091962324,
      "loss": 5.1681,
      "mean_token_accuracy": 0.20488449037075043,
      "num_tokens": 18060163.0,
      "step": 7875
    },
    {
      "entropy": 5.36887936592102,
      "epoch": 0.7569644572526417,
      "grad_norm": 0.875,
      "learning_rate": 0.0004950741773268788,
      "loss": 5.1173,
      "mean_token_accuracy": 0.2073910653591156,
      "num_tokens": 18071628.0,
      "step": 7880
    },
    {
      "entropy": 5.252798938751221,
      "epoch": 0.7574447646493756,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004950670403480562,
      "loss": 5.0227,
      "mean_token_accuracy": 0.21277171969413758,
      "num_tokens": 18082979.0,
      "step": 7885
    },
    {
      "entropy": 5.323721408843994,
      "epoch": 0.7579250720461095,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004950598982599306,
      "loss": 5.0742,
      "mean_token_accuracy": 0.21375603824853898,
      "num_tokens": 18093889.0,
      "step": 7890
    },
    {
      "entropy": 5.307314872741699,
      "epoch": 0.7584053794428435,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004950527510626675,
      "loss": 5.1215,
      "mean_token_accuracy": 0.20746706873178483,
      "num_tokens": 18105559.0,
      "step": 7895
    },
    {
      "entropy": 5.351215410232544,
      "epoch": 0.7588856868395774,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004950455987564329,
      "loss": 5.2053,
      "mean_token_accuracy": 0.2072554424405098,
      "num_tokens": 18116316.0,
      "step": 7900
    },
    {
      "entropy": 5.393552827835083,
      "epoch": 0.7593659942363112,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004950384413413931,
      "loss": 5.0928,
      "mean_token_accuracy": 0.21228759735822678,
      "num_tokens": 18126851.0,
      "step": 7905
    },
    {
      "entropy": 5.39699501991272,
      "epoch": 0.7598463016330451,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004950312788177139,
      "loss": 5.1745,
      "mean_token_accuracy": 0.2036636143922806,
      "num_tokens": 18139571.0,
      "step": 7910
    },
    {
      "entropy": 5.425048732757569,
      "epoch": 0.7603266090297791,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004950241111855618,
      "loss": 5.162,
      "mean_token_accuracy": 0.20627158880233765,
      "num_tokens": 18150680.0,
      "step": 7915
    },
    {
      "entropy": 5.368895626068115,
      "epoch": 0.760806916426513,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004950169384451031,
      "loss": 5.1115,
      "mean_token_accuracy": 0.21384803652763368,
      "num_tokens": 18161911.0,
      "step": 7920
    },
    {
      "entropy": 5.361880826950073,
      "epoch": 0.7612872238232469,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004950097605965045,
      "loss": 5.0985,
      "mean_token_accuracy": 0.2134673923254013,
      "num_tokens": 18172714.0,
      "step": 7925
    },
    {
      "entropy": 5.34326319694519,
      "epoch": 0.7617675312199808,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004950025776399326,
      "loss": 5.0878,
      "mean_token_accuracy": 0.21684494465589524,
      "num_tokens": 18184250.0,
      "step": 7930
    },
    {
      "entropy": 5.411981201171875,
      "epoch": 0.7622478386167147,
      "grad_norm": 0.859375,
      "learning_rate": 0.000494995389575554,
      "loss": 5.2062,
      "mean_token_accuracy": 0.20142002552747726,
      "num_tokens": 18195299.0,
      "step": 7935
    },
    {
      "entropy": 5.427820301055908,
      "epoch": 0.7627281460134486,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004949881964035357,
      "loss": 5.2638,
      "mean_token_accuracy": 0.2053232327103615,
      "num_tokens": 18206863.0,
      "step": 7940
    },
    {
      "entropy": 5.404353332519531,
      "epoch": 0.7632084534101825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004949809981240448,
      "loss": 5.1849,
      "mean_token_accuracy": 0.20650373995304108,
      "num_tokens": 18219643.0,
      "step": 7945
    },
    {
      "entropy": 5.327841711044312,
      "epoch": 0.7636887608069164,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004949737947372483,
      "loss": 5.0344,
      "mean_token_accuracy": 0.21734251081943512,
      "num_tokens": 18230461.0,
      "step": 7950
    },
    {
      "entropy": 5.325031423568726,
      "epoch": 0.7641690682036504,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004949665862433134,
      "loss": 5.1711,
      "mean_token_accuracy": 0.20416854768991471,
      "num_tokens": 18243768.0,
      "step": 7955
    },
    {
      "entropy": 5.375663566589355,
      "epoch": 0.7646493756003843,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004949593726424077,
      "loss": 5.0716,
      "mean_token_accuracy": 0.21150785088539123,
      "num_tokens": 18255322.0,
      "step": 7960
    },
    {
      "entropy": 5.341422843933105,
      "epoch": 0.7651296829971181,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004949521539346985,
      "loss": 5.1222,
      "mean_token_accuracy": 0.21221319288015367,
      "num_tokens": 18267262.0,
      "step": 7965
    },
    {
      "entropy": 5.296201753616333,
      "epoch": 0.765609990393852,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004949449301203533,
      "loss": 5.0249,
      "mean_token_accuracy": 0.20996281653642654,
      "num_tokens": 18277332.0,
      "step": 7970
    },
    {
      "entropy": 5.32009482383728,
      "epoch": 0.766090297790586,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004949377011995399,
      "loss": 5.1597,
      "mean_token_accuracy": 0.2102154850959778,
      "num_tokens": 18289547.0,
      "step": 7975
    },
    {
      "entropy": 5.3457835674285885,
      "epoch": 0.7665706051873199,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004949304671724263,
      "loss": 5.061,
      "mean_token_accuracy": 0.21181625723838807,
      "num_tokens": 18302097.0,
      "step": 7980
    },
    {
      "entropy": 5.363691997528076,
      "epoch": 0.7670509125840538,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004949232280391802,
      "loss": 5.1333,
      "mean_token_accuracy": 0.20804425030946733,
      "num_tokens": 18314151.0,
      "step": 7985
    },
    {
      "entropy": 5.287468576431275,
      "epoch": 0.7675312199807877,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004949159837999698,
      "loss": 5.0781,
      "mean_token_accuracy": 0.20622672289609909,
      "num_tokens": 18326085.0,
      "step": 7990
    },
    {
      "entropy": 5.312514114379883,
      "epoch": 0.7680115273775217,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004949087344549633,
      "loss": 5.0661,
      "mean_token_accuracy": 0.21037355363368987,
      "num_tokens": 18338158.0,
      "step": 7995
    },
    {
      "entropy": 5.364059734344482,
      "epoch": 0.7684918347742555,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000494901480004329,
      "loss": 5.0639,
      "mean_token_accuracy": 0.21409421414136887,
      "num_tokens": 18349418.0,
      "step": 8000
    },
    {
      "entropy": 5.344389581680298,
      "epoch": 0.7689721421709894,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004948942204482351,
      "loss": 5.0973,
      "mean_token_accuracy": 0.20933580696582793,
      "num_tokens": 18361964.0,
      "step": 8005
    },
    {
      "entropy": 5.274578428268432,
      "epoch": 0.7694524495677233,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004948869557868506,
      "loss": 4.9943,
      "mean_token_accuracy": 0.22064043432474137,
      "num_tokens": 18373108.0,
      "step": 8010
    },
    {
      "entropy": 5.237949657440185,
      "epoch": 0.7699327569644573,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004948796860203439,
      "loss": 5.0063,
      "mean_token_accuracy": 0.2173450067639351,
      "num_tokens": 18385310.0,
      "step": 8015
    },
    {
      "entropy": 5.365011930465698,
      "epoch": 0.7704130643611912,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004948724111488838,
      "loss": 5.2042,
      "mean_token_accuracy": 0.20785281360149382,
      "num_tokens": 18396132.0,
      "step": 8020
    },
    {
      "entropy": 5.495370006561279,
      "epoch": 0.770893371757925,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004948651311726391,
      "loss": 5.2902,
      "mean_token_accuracy": 0.20274072885513306,
      "num_tokens": 18409930.0,
      "step": 8025
    },
    {
      "entropy": 5.417250537872315,
      "epoch": 0.7713736791546589,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004948578460917789,
      "loss": 5.1818,
      "mean_token_accuracy": 0.20224556624889373,
      "num_tokens": 18421204.0,
      "step": 8030
    },
    {
      "entropy": 5.347210025787353,
      "epoch": 0.7718539865513929,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004948505559064725,
      "loss": 5.0746,
      "mean_token_accuracy": 0.21489468216896057,
      "num_tokens": 18433194.0,
      "step": 8035
    },
    {
      "entropy": 5.293833208084107,
      "epoch": 0.7723342939481268,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004948432606168889,
      "loss": 5.0411,
      "mean_token_accuracy": 0.21716948300600053,
      "num_tokens": 18445282.0,
      "step": 8040
    },
    {
      "entropy": 5.332662916183471,
      "epoch": 0.7728146013448607,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004948359602231976,
      "loss": 5.0722,
      "mean_token_accuracy": 0.21722075641155242,
      "num_tokens": 18456264.0,
      "step": 8045
    },
    {
      "entropy": 5.2765247344970705,
      "epoch": 0.7732949087415946,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004948286547255681,
      "loss": 5.02,
      "mean_token_accuracy": 0.21654772758483887,
      "num_tokens": 18467573.0,
      "step": 8050
    },
    {
      "entropy": 5.296917343139649,
      "epoch": 0.7737752161383286,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00049482134412417,
      "loss": 4.9953,
      "mean_token_accuracy": 0.22097841203212737,
      "num_tokens": 18478333.0,
      "step": 8055
    },
    {
      "entropy": 5.27380485534668,
      "epoch": 0.7742555235350624,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000494814028419173,
      "loss": 5.0171,
      "mean_token_accuracy": 0.21664547473192214,
      "num_tokens": 18489106.0,
      "step": 8060
    },
    {
      "entropy": 5.3441237449646,
      "epoch": 0.7747358309317963,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000494806707610747,
      "loss": 5.1444,
      "mean_token_accuracy": 0.208481964468956,
      "num_tokens": 18500090.0,
      "step": 8065
    },
    {
      "entropy": 5.353561973571777,
      "epoch": 0.7752161383285303,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000494799381699062,
      "loss": 5.108,
      "mean_token_accuracy": 0.21635482758283614,
      "num_tokens": 18510305.0,
      "step": 8070
    },
    {
      "entropy": 5.343230533599853,
      "epoch": 0.7756964457252642,
      "grad_norm": 0.875,
      "learning_rate": 0.0004947920506842879,
      "loss": 5.1122,
      "mean_token_accuracy": 0.20923472940921783,
      "num_tokens": 18523048.0,
      "step": 8075
    },
    {
      "entropy": 5.284622097015381,
      "epoch": 0.7761767531219981,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004947847145665951,
      "loss": 5.0249,
      "mean_token_accuracy": 0.21444708853960037,
      "num_tokens": 18534145.0,
      "step": 8080
    },
    {
      "entropy": 5.320905447006226,
      "epoch": 0.776657060518732,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004947773733461539,
      "loss": 5.0754,
      "mean_token_accuracy": 0.2129151463508606,
      "num_tokens": 18545045.0,
      "step": 8085
    },
    {
      "entropy": 5.320458459854126,
      "epoch": 0.777137367915466,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004947700270231347,
      "loss": 5.0626,
      "mean_token_accuracy": 0.21323574781417848,
      "num_tokens": 18557531.0,
      "step": 8090
    },
    {
      "entropy": 5.346653175354004,
      "epoch": 0.7776176753121998,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004947626755977079,
      "loss": 5.1103,
      "mean_token_accuracy": 0.20757609605789185,
      "num_tokens": 18569127.0,
      "step": 8095
    },
    {
      "entropy": 5.376105451583863,
      "epoch": 0.7780979827089337,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004947553190700444,
      "loss": 5.1298,
      "mean_token_accuracy": 0.21466040015220642,
      "num_tokens": 18580606.0,
      "step": 8100
    },
    {
      "entropy": 5.413986730575561,
      "epoch": 0.7785782901056676,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000494747957440315,
      "loss": 5.238,
      "mean_token_accuracy": 0.20085193514823912,
      "num_tokens": 18592330.0,
      "step": 8105
    },
    {
      "entropy": 5.409378528594971,
      "epoch": 0.7790585975024016,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004947405907086905,
      "loss": 5.2095,
      "mean_token_accuracy": 0.20601927787065505,
      "num_tokens": 18604721.0,
      "step": 8110
    },
    {
      "entropy": 5.336650800704956,
      "epoch": 0.7795389048991355,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004947332188753419,
      "loss": 5.0874,
      "mean_token_accuracy": 0.2152855157852173,
      "num_tokens": 18616814.0,
      "step": 8115
    },
    {
      "entropy": 5.3368322372436525,
      "epoch": 0.7800192122958693,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004947258419404405,
      "loss": 5.0934,
      "mean_token_accuracy": 0.2123816654086113,
      "num_tokens": 18628224.0,
      "step": 8120
    },
    {
      "entropy": 5.339434003829956,
      "epoch": 0.7804995196926032,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004947184599041576,
      "loss": 5.1329,
      "mean_token_accuracy": 0.2134467914700508,
      "num_tokens": 18639777.0,
      "step": 8125
    },
    {
      "entropy": 5.280817317962646,
      "epoch": 0.7809798270893372,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004947110727666644,
      "loss": 5.0568,
      "mean_token_accuracy": 0.2116105005145073,
      "num_tokens": 18651044.0,
      "step": 8130
    },
    {
      "entropy": 5.358986330032349,
      "epoch": 0.7814601344860711,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004947036805281325,
      "loss": 5.1128,
      "mean_token_accuracy": 0.2121086075901985,
      "num_tokens": 18663142.0,
      "step": 8135
    },
    {
      "entropy": 5.436714458465576,
      "epoch": 0.781940441882805,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004946962831887336,
      "loss": 5.193,
      "mean_token_accuracy": 0.20681314170360565,
      "num_tokens": 18674079.0,
      "step": 8140
    },
    {
      "entropy": 5.346633863449097,
      "epoch": 0.7824207492795389,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004946888807486393,
      "loss": 5.1711,
      "mean_token_accuracy": 0.2068258762359619,
      "num_tokens": 18685744.0,
      "step": 8145
    },
    {
      "entropy": 5.2618049621582035,
      "epoch": 0.7829010566762729,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004946814732080214,
      "loss": 5.057,
      "mean_token_accuracy": 0.21336293667554856,
      "num_tokens": 18697049.0,
      "step": 8150
    },
    {
      "entropy": 5.298804378509521,
      "epoch": 0.7833813640730067,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004946740605670523,
      "loss": 5.0683,
      "mean_token_accuracy": 0.21712972670793534,
      "num_tokens": 18708765.0,
      "step": 8155
    },
    {
      "entropy": 5.357657814025879,
      "epoch": 0.7838616714697406,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004946666428259037,
      "loss": 5.0556,
      "mean_token_accuracy": 0.21680933535099028,
      "num_tokens": 18719819.0,
      "step": 8160
    },
    {
      "entropy": 5.390479183197021,
      "epoch": 0.7843419788664745,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004946592199847478,
      "loss": 5.1186,
      "mean_token_accuracy": 0.2056015208363533,
      "num_tokens": 18730668.0,
      "step": 8165
    },
    {
      "entropy": 5.240038919448852,
      "epoch": 0.7848222862632085,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004946517920437571,
      "loss": 4.9265,
      "mean_token_accuracy": 0.21651222854852675,
      "num_tokens": 18741991.0,
      "step": 8170
    },
    {
      "entropy": 5.334393453598023,
      "epoch": 0.7853025936599424,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004946443590031041,
      "loss": 5.1446,
      "mean_token_accuracy": 0.2118510514497757,
      "num_tokens": 18753919.0,
      "step": 8175
    },
    {
      "entropy": 5.362671709060669,
      "epoch": 0.7857829010566763,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004946369208629613,
      "loss": 5.0513,
      "mean_token_accuracy": 0.21652378141880035,
      "num_tokens": 18764342.0,
      "step": 8180
    },
    {
      "entropy": 5.259613752365112,
      "epoch": 0.7862632084534101,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004946294776235013,
      "loss": 5.0574,
      "mean_token_accuracy": 0.21248440444469452,
      "num_tokens": 18776044.0,
      "step": 8185
    },
    {
      "entropy": 5.386792850494385,
      "epoch": 0.7867435158501441,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004946220292848971,
      "loss": 5.1354,
      "mean_token_accuracy": 0.21177714318037033,
      "num_tokens": 18787354.0,
      "step": 8190
    },
    {
      "entropy": 5.328051328659058,
      "epoch": 0.787223823246878,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004946145758473214,
      "loss": 5.0435,
      "mean_token_accuracy": 0.20805221796035767,
      "num_tokens": 18797845.0,
      "step": 8195
    },
    {
      "entropy": 5.388686037063598,
      "epoch": 0.7877041306436119,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004946071173109475,
      "loss": 5.1253,
      "mean_token_accuracy": 0.21109261512756347,
      "num_tokens": 18809253.0,
      "step": 8200
    },
    {
      "entropy": 5.31988844871521,
      "epoch": 0.7881844380403458,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004945996536759484,
      "loss": 4.9841,
      "mean_token_accuracy": 0.22203437983989716,
      "num_tokens": 18819768.0,
      "step": 8205
    },
    {
      "entropy": 5.420179939270019,
      "epoch": 0.7886647454370798,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004945921849424974,
      "loss": 5.2413,
      "mean_token_accuracy": 0.203597816824913,
      "num_tokens": 18831151.0,
      "step": 8210
    },
    {
      "entropy": 5.346021032333374,
      "epoch": 0.7891450528338136,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004945847111107679,
      "loss": 5.0267,
      "mean_token_accuracy": 0.21523899137973784,
      "num_tokens": 18842133.0,
      "step": 8215
    },
    {
      "entropy": 5.344703817367554,
      "epoch": 0.7896253602305475,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004945772321809334,
      "loss": 5.1699,
      "mean_token_accuracy": 0.20583187490701677,
      "num_tokens": 18853295.0,
      "step": 8220
    },
    {
      "entropy": 5.327058267593384,
      "epoch": 0.7901056676272814,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004945697481531677,
      "loss": 5.1224,
      "mean_token_accuracy": 0.2152056321501732,
      "num_tokens": 18865802.0,
      "step": 8225
    },
    {
      "entropy": 5.34009199142456,
      "epoch": 0.7905859750240154,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004945622590276443,
      "loss": 5.0398,
      "mean_token_accuracy": 0.21951935440301895,
      "num_tokens": 18877693.0,
      "step": 8230
    },
    {
      "entropy": 5.331456184387207,
      "epoch": 0.7910662824207493,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004945547648045373,
      "loss": 5.0907,
      "mean_token_accuracy": 0.216140778362751,
      "num_tokens": 18888549.0,
      "step": 8235
    },
    {
      "entropy": 5.311357641220093,
      "epoch": 0.7915465898174832,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004945472654840206,
      "loss": 5.031,
      "mean_token_accuracy": 0.22045050114393233,
      "num_tokens": 18899132.0,
      "step": 8240
    },
    {
      "entropy": 5.214868497848511,
      "epoch": 0.7920268972142172,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004945397610662683,
      "loss": 5.0456,
      "mean_token_accuracy": 0.21102908104658127,
      "num_tokens": 18911774.0,
      "step": 8245
    },
    {
      "entropy": 5.359712934494018,
      "epoch": 0.792507204610951,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004945322515514547,
      "loss": 5.1095,
      "mean_token_accuracy": 0.21240750849246978,
      "num_tokens": 18922806.0,
      "step": 8250
    },
    {
      "entropy": 5.3473938465118405,
      "epoch": 0.7929875120076849,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000494524736939754,
      "loss": 5.0836,
      "mean_token_accuracy": 0.2183587819337845,
      "num_tokens": 18934861.0,
      "step": 8255
    },
    {
      "entropy": 5.3650988101959225,
      "epoch": 0.7934678194044188,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004945172172313408,
      "loss": 5.066,
      "mean_token_accuracy": 0.2158061996102333,
      "num_tokens": 18944269.0,
      "step": 8260
    },
    {
      "entropy": 5.311491918563843,
      "epoch": 0.7939481268011528,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004945096924263896,
      "loss": 5.1103,
      "mean_token_accuracy": 0.20884205251932145,
      "num_tokens": 18957197.0,
      "step": 8265
    },
    {
      "entropy": 5.361815977096557,
      "epoch": 0.7944284341978867,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004945021625250753,
      "loss": 5.1107,
      "mean_token_accuracy": 0.20595426708459855,
      "num_tokens": 18968254.0,
      "step": 8270
    },
    {
      "entropy": 5.290348529815674,
      "epoch": 0.7949087415946205,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004944946275275724,
      "loss": 5.0172,
      "mean_token_accuracy": 0.2212981402873993,
      "num_tokens": 18979372.0,
      "step": 8275
    },
    {
      "entropy": 5.313999795913697,
      "epoch": 0.7953890489913544,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004944870874340561,
      "loss": 5.0305,
      "mean_token_accuracy": 0.21565368622541428,
      "num_tokens": 18991075.0,
      "step": 8280
    },
    {
      "entropy": 5.274328660964966,
      "epoch": 0.7958693563880884,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004944795422447013,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21833001971244811,
      "num_tokens": 19002324.0,
      "step": 8285
    },
    {
      "entropy": 5.35863938331604,
      "epoch": 0.7963496637848223,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004944719919596835,
      "loss": 5.1438,
      "mean_token_accuracy": 0.209305140376091,
      "num_tokens": 19014406.0,
      "step": 8290
    },
    {
      "entropy": 5.293546295166015,
      "epoch": 0.7968299711815562,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004944644365791776,
      "loss": 5.0238,
      "mean_token_accuracy": 0.2161477640271187,
      "num_tokens": 19025984.0,
      "step": 8295
    },
    {
      "entropy": 5.361144304275513,
      "epoch": 0.7973102785782901,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000494456876103359,
      "loss": 5.1329,
      "mean_token_accuracy": 0.21390934139490128,
      "num_tokens": 19036189.0,
      "step": 8300
    },
    {
      "entropy": 5.36251425743103,
      "epoch": 0.7977905859750241,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004944493105324035,
      "loss": 5.1217,
      "mean_token_accuracy": 0.20638122409582138,
      "num_tokens": 19047587.0,
      "step": 8305
    },
    {
      "entropy": 5.362772703170776,
      "epoch": 0.7982708933717579,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004944417398664866,
      "loss": 5.0909,
      "mean_token_accuracy": 0.2113674685359001,
      "num_tokens": 19058467.0,
      "step": 8310
    },
    {
      "entropy": 5.344825744628906,
      "epoch": 0.7987512007684918,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004944341641057843,
      "loss": 5.1682,
      "mean_token_accuracy": 0.20513344258069993,
      "num_tokens": 19070235.0,
      "step": 8315
    },
    {
      "entropy": 5.384693574905396,
      "epoch": 0.7992315081652257,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004944265832504721,
      "loss": 5.1256,
      "mean_token_accuracy": 0.20910855382680893,
      "num_tokens": 19082005.0,
      "step": 8320
    },
    {
      "entropy": 5.387254333496093,
      "epoch": 0.7997118155619597,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004944189973007262,
      "loss": 5.137,
      "mean_token_accuracy": 0.20899821519851686,
      "num_tokens": 19092922.0,
      "step": 8325
    },
    {
      "entropy": 5.358731603622436,
      "epoch": 0.8001921229586936,
      "grad_norm": 0.875,
      "learning_rate": 0.0004944114062567229,
      "loss": 5.1403,
      "mean_token_accuracy": 0.20726601481437684,
      "num_tokens": 19104832.0,
      "step": 8330
    },
    {
      "entropy": 5.353858566284179,
      "epoch": 0.8006724303554275,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004944038101186381,
      "loss": 5.1676,
      "mean_token_accuracy": 0.20989611446857454,
      "num_tokens": 19116261.0,
      "step": 8335
    },
    {
      "entropy": 5.385284328460694,
      "epoch": 0.8011527377521613,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004943962088866483,
      "loss": 5.1465,
      "mean_token_accuracy": 0.208486644923687,
      "num_tokens": 19127195.0,
      "step": 8340
    },
    {
      "entropy": 5.419732570648193,
      "epoch": 0.8016330451488953,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004943886025609301,
      "loss": 5.1791,
      "mean_token_accuracy": 0.2045424312353134,
      "num_tokens": 19138164.0,
      "step": 8345
    },
    {
      "entropy": 5.360116863250733,
      "epoch": 0.8021133525456292,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00049438099114166,
      "loss": 5.1321,
      "mean_token_accuracy": 0.20329462736845016,
      "num_tokens": 19149153.0,
      "step": 8350
    },
    {
      "entropy": 5.361733770370483,
      "epoch": 0.8025936599423631,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004943733746290147,
      "loss": 5.166,
      "mean_token_accuracy": 0.20284549593925477,
      "num_tokens": 19161023.0,
      "step": 8355
    },
    {
      "entropy": 5.287382888793945,
      "epoch": 0.803073967339097,
      "grad_norm": 1.0625,
      "learning_rate": 0.000494365753023171,
      "loss": 5.0115,
      "mean_token_accuracy": 0.2190767154097557,
      "num_tokens": 19172466.0,
      "step": 8360
    },
    {
      "entropy": 5.356825351715088,
      "epoch": 0.803554274735831,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004943581263243059,
      "loss": 5.0993,
      "mean_token_accuracy": 0.2144365444779396,
      "num_tokens": 19183684.0,
      "step": 8365
    },
    {
      "entropy": 5.272908306121826,
      "epoch": 0.8040345821325648,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004943504945325965,
      "loss": 5.0071,
      "mean_token_accuracy": 0.21848383843898772,
      "num_tokens": 19194479.0,
      "step": 8370
    },
    {
      "entropy": 5.365400075912476,
      "epoch": 0.8045148895292987,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004943428576482198,
      "loss": 5.0829,
      "mean_token_accuracy": 0.20531211644411088,
      "num_tokens": 19206323.0,
      "step": 8375
    },
    {
      "entropy": 5.431076002120972,
      "epoch": 0.8049951969260326,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004943352156713535,
      "loss": 5.1671,
      "mean_token_accuracy": 0.20022738873958587,
      "num_tokens": 19218849.0,
      "step": 8380
    },
    {
      "entropy": 5.293436479568482,
      "epoch": 0.8054755043227666,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004943275686021747,
      "loss": 4.9943,
      "mean_token_accuracy": 0.21352229863405228,
      "num_tokens": 19229603.0,
      "step": 8385
    },
    {
      "entropy": 5.34028811454773,
      "epoch": 0.8059558117195005,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000494319916440861,
      "loss": 5.1635,
      "mean_token_accuracy": 0.2062859371304512,
      "num_tokens": 19241318.0,
      "step": 8390
    },
    {
      "entropy": 5.355140495300293,
      "epoch": 0.8064361191162344,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004943122591875901,
      "loss": 5.051,
      "mean_token_accuracy": 0.2132823273539543,
      "num_tokens": 19252640.0,
      "step": 8395
    },
    {
      "entropy": 5.3117975234985355,
      "epoch": 0.8069164265129684,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004943045968425398,
      "loss": 5.0483,
      "mean_token_accuracy": 0.21493369936943055,
      "num_tokens": 19262971.0,
      "step": 8400
    },
    {
      "entropy": 5.378424739837646,
      "epoch": 0.8073967339097022,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004942969294058878,
      "loss": 5.1369,
      "mean_token_accuracy": 0.2132812350988388,
      "num_tokens": 19274426.0,
      "step": 8405
    },
    {
      "entropy": 5.337623071670532,
      "epoch": 0.8078770413064361,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004942892568778125,
      "loss": 5.0983,
      "mean_token_accuracy": 0.21478380113840104,
      "num_tokens": 19286806.0,
      "step": 8410
    },
    {
      "entropy": 5.286077880859375,
      "epoch": 0.80835734870317,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004942815792584917,
      "loss": 5.0645,
      "mean_token_accuracy": 0.21373484581708907,
      "num_tokens": 19297997.0,
      "step": 8415
    },
    {
      "entropy": 5.310636758804321,
      "epoch": 0.808837656099904,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004942738965481038,
      "loss": 5.0421,
      "mean_token_accuracy": 0.21546782851219176,
      "num_tokens": 19309789.0,
      "step": 8420
    },
    {
      "entropy": 5.289216232299805,
      "epoch": 0.8093179634966379,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004942662087468272,
      "loss": 5.0663,
      "mean_token_accuracy": 0.21881039291620255,
      "num_tokens": 19320688.0,
      "step": 8425
    },
    {
      "entropy": 5.311178970336914,
      "epoch": 0.8097982708933718,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004942585158548402,
      "loss": 5.0989,
      "mean_token_accuracy": 0.21207071393728255,
      "num_tokens": 19331670.0,
      "step": 8430
    },
    {
      "entropy": 5.360179996490478,
      "epoch": 0.8102785782901056,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004942508178723214,
      "loss": 5.0474,
      "mean_token_accuracy": 0.21228671967983245,
      "num_tokens": 19343578.0,
      "step": 8435
    },
    {
      "entropy": 5.34991226196289,
      "epoch": 0.8107588856868396,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004942431147994499,
      "loss": 5.1017,
      "mean_token_accuracy": 0.2092154800891876,
      "num_tokens": 19354875.0,
      "step": 8440
    },
    {
      "entropy": 5.283786773681641,
      "epoch": 0.8112391930835735,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004942354066364042,
      "loss": 5.0501,
      "mean_token_accuracy": 0.21562657952308656,
      "num_tokens": 19366353.0,
      "step": 8445
    },
    {
      "entropy": 5.2197428226470945,
      "epoch": 0.8117195004803074,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004942276933833634,
      "loss": 4.983,
      "mean_token_accuracy": 0.22176084220409392,
      "num_tokens": 19377534.0,
      "step": 8450
    },
    {
      "entropy": 5.3212579727172855,
      "epoch": 0.8121998078770413,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004942199750405064,
      "loss": 5.1215,
      "mean_token_accuracy": 0.20401201248168946,
      "num_tokens": 19388188.0,
      "step": 8455
    },
    {
      "entropy": 5.222447204589844,
      "epoch": 0.8126801152737753,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004942122516080127,
      "loss": 4.9575,
      "mean_token_accuracy": 0.22281467020511628,
      "num_tokens": 19399910.0,
      "step": 8460
    },
    {
      "entropy": 5.3330730438232425,
      "epoch": 0.8131604226705091,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004942045230860614,
      "loss": 5.1082,
      "mean_token_accuracy": 0.21519561409950255,
      "num_tokens": 19411715.0,
      "step": 8465
    },
    {
      "entropy": 5.42485237121582,
      "epoch": 0.813640730067243,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004941967894748319,
      "loss": 5.2046,
      "mean_token_accuracy": 0.19849235266447068,
      "num_tokens": 19423275.0,
      "step": 8470
    },
    {
      "entropy": 5.3768164157867435,
      "epoch": 0.8141210374639769,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004941890507745039,
      "loss": 5.1234,
      "mean_token_accuracy": 0.20622891187667847,
      "num_tokens": 19436035.0,
      "step": 8475
    },
    {
      "entropy": 5.261404514312744,
      "epoch": 0.8146013448607109,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004941813069852569,
      "loss": 5.0867,
      "mean_token_accuracy": 0.20943611413240432,
      "num_tokens": 19447755.0,
      "step": 8480
    },
    {
      "entropy": 5.351346874237061,
      "epoch": 0.8150816522574448,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004941735581072708,
      "loss": 5.0542,
      "mean_token_accuracy": 0.2174903616309166,
      "num_tokens": 19459044.0,
      "step": 8485
    },
    {
      "entropy": 5.320174598693848,
      "epoch": 0.8155619596541787,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004941658041407255,
      "loss": 5.0355,
      "mean_token_accuracy": 0.21346810907125474,
      "num_tokens": 19471486.0,
      "step": 8490
    },
    {
      "entropy": 5.237428140640259,
      "epoch": 0.8160422670509125,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000494158045085801,
      "loss": 4.9733,
      "mean_token_accuracy": 0.22339199632406234,
      "num_tokens": 19482428.0,
      "step": 8495
    },
    {
      "entropy": 5.329161167144775,
      "epoch": 0.8165225744476465,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004941502809426776,
      "loss": 5.1393,
      "mean_token_accuracy": 0.20986351370811462,
      "num_tokens": 19494616.0,
      "step": 8500
    },
    {
      "entropy": 5.336520338058472,
      "epoch": 0.8170028818443804,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004941425117115354,
      "loss": 5.0343,
      "mean_token_accuracy": 0.21424599438905717,
      "num_tokens": 19504953.0,
      "step": 8505
    },
    {
      "entropy": 5.3431257724761965,
      "epoch": 0.8174831892411143,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004941347373925547,
      "loss": 5.1247,
      "mean_token_accuracy": 0.21505391895771026,
      "num_tokens": 19517408.0,
      "step": 8510
    },
    {
      "entropy": 5.2607475280761715,
      "epoch": 0.8179634966378482,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004941269579859161,
      "loss": 5.0459,
      "mean_token_accuracy": 0.22004921734333038,
      "num_tokens": 19529190.0,
      "step": 8515
    },
    {
      "entropy": 5.31341233253479,
      "epoch": 0.8184438040345822,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004941191734918002,
      "loss": 5.0882,
      "mean_token_accuracy": 0.20821243971586229,
      "num_tokens": 19540278.0,
      "step": 8520
    },
    {
      "entropy": 5.31887526512146,
      "epoch": 0.818924111431316,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000494111383910388,
      "loss": 5.065,
      "mean_token_accuracy": 0.21409234553575515,
      "num_tokens": 19551224.0,
      "step": 8525
    },
    {
      "entropy": 5.388135528564453,
      "epoch": 0.8194044188280499,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004941035892418597,
      "loss": 5.0846,
      "mean_token_accuracy": 0.2096189007163048,
      "num_tokens": 19564386.0,
      "step": 8530
    },
    {
      "entropy": 5.348142671585083,
      "epoch": 0.8198847262247838,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004940957894863968,
      "loss": 5.0768,
      "mean_token_accuracy": 0.21463494449853898,
      "num_tokens": 19576170.0,
      "step": 8535
    },
    {
      "entropy": 5.300370883941651,
      "epoch": 0.8203650336215178,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004940879846441804,
      "loss": 5.0615,
      "mean_token_accuracy": 0.21069518327713013,
      "num_tokens": 19587220.0,
      "step": 8540
    },
    {
      "entropy": 5.297258901596069,
      "epoch": 0.8208453410182517,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004940801747153914,
      "loss": 5.0642,
      "mean_token_accuracy": 0.21124200373888016,
      "num_tokens": 19598649.0,
      "step": 8545
    },
    {
      "entropy": 5.337055778503418,
      "epoch": 0.8213256484149856,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004940723597002113,
      "loss": 5.1286,
      "mean_token_accuracy": 0.20770150423049927,
      "num_tokens": 19610243.0,
      "step": 8550
    },
    {
      "entropy": 5.345717525482177,
      "epoch": 0.8218059558117194,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004940645395988216,
      "loss": 5.1247,
      "mean_token_accuracy": 0.20248719155788422,
      "num_tokens": 19621467.0,
      "step": 8555
    },
    {
      "entropy": 5.408337879180908,
      "epoch": 0.8222862632084534,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004940567144114036,
      "loss": 5.1957,
      "mean_token_accuracy": 0.2011337772011757,
      "num_tokens": 19633367.0,
      "step": 8560
    },
    {
      "entropy": 5.281939888000489,
      "epoch": 0.8227665706051873,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004940488841381393,
      "loss": 5.055,
      "mean_token_accuracy": 0.21791253834962845,
      "num_tokens": 19643144.0,
      "step": 8565
    },
    {
      "entropy": 5.310028409957885,
      "epoch": 0.8232468780019212,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004940410487792103,
      "loss": 5.1479,
      "mean_token_accuracy": 0.20937480330467223,
      "num_tokens": 19654501.0,
      "step": 8570
    },
    {
      "entropy": 5.369356441497803,
      "epoch": 0.8237271853986552,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004940332083347986,
      "loss": 5.0959,
      "mean_token_accuracy": 0.2079701691865921,
      "num_tokens": 19665382.0,
      "step": 8575
    },
    {
      "entropy": 5.303474950790405,
      "epoch": 0.8242074927953891,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004940253628050861,
      "loss": 4.943,
      "mean_token_accuracy": 0.22071055173873902,
      "num_tokens": 19677222.0,
      "step": 8580
    },
    {
      "entropy": 5.245962619781494,
      "epoch": 0.824687800192123,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004940175121902552,
      "loss": 5.0415,
      "mean_token_accuracy": 0.21321824938058853,
      "num_tokens": 19688550.0,
      "step": 8585
    },
    {
      "entropy": 5.379223823547363,
      "epoch": 0.8251681075888568,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000494009656490488,
      "loss": 5.142,
      "mean_token_accuracy": 0.20742358565330504,
      "num_tokens": 19700125.0,
      "step": 8590
    },
    {
      "entropy": 5.384274339675903,
      "epoch": 0.8256484149855908,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004940017957059668,
      "loss": 5.1513,
      "mean_token_accuracy": 0.20902712941169738,
      "num_tokens": 19711969.0,
      "step": 8595
    },
    {
      "entropy": 5.299255466461181,
      "epoch": 0.8261287223823247,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004939939298368742,
      "loss": 5.0994,
      "mean_token_accuracy": 0.21238535046577453,
      "num_tokens": 19723813.0,
      "step": 8600
    },
    {
      "entropy": 5.360401582717896,
      "epoch": 0.8266090297790586,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004939860588833929,
      "loss": 5.1268,
      "mean_token_accuracy": 0.20373987406492233,
      "num_tokens": 19735539.0,
      "step": 8605
    },
    {
      "entropy": 5.457154989242554,
      "epoch": 0.8270893371757925,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004939781828457055,
      "loss": 5.199,
      "mean_token_accuracy": 0.20703657418489457,
      "num_tokens": 19747106.0,
      "step": 8610
    },
    {
      "entropy": 5.3150077819824215,
      "epoch": 0.8275696445725265,
      "grad_norm": 0.890625,
      "learning_rate": 0.000493970301723995,
      "loss": 5.0938,
      "mean_token_accuracy": 0.22241331785917282,
      "num_tokens": 19760156.0,
      "step": 8615
    },
    {
      "entropy": 5.269891834259033,
      "epoch": 0.8280499519692603,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004939624155184443,
      "loss": 4.9966,
      "mean_token_accuracy": 0.21598958373069763,
      "num_tokens": 19771256.0,
      "step": 8620
    },
    {
      "entropy": 5.323619985580445,
      "epoch": 0.8285302593659942,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939545242292365,
      "loss": 5.0125,
      "mean_token_accuracy": 0.21737516075372695,
      "num_tokens": 19781332.0,
      "step": 8625
    },
    {
      "entropy": 5.35371150970459,
      "epoch": 0.8290105667627281,
      "grad_norm": 0.875,
      "learning_rate": 0.0004939466278565547,
      "loss": 5.1758,
      "mean_token_accuracy": 0.20606835633516313,
      "num_tokens": 19793573.0,
      "step": 8630
    },
    {
      "entropy": 5.2834676742553714,
      "epoch": 0.8294908741594621,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939387264005825,
      "loss": 5.0712,
      "mean_token_accuracy": 0.2189349576830864,
      "num_tokens": 19803421.0,
      "step": 8635
    },
    {
      "entropy": 5.22930212020874,
      "epoch": 0.829971181556196,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004939308198615031,
      "loss": 4.9402,
      "mean_token_accuracy": 0.22577660232782365,
      "num_tokens": 19814440.0,
      "step": 8640
    },
    {
      "entropy": 5.253708362579346,
      "epoch": 0.8304514889529299,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004939229082395001,
      "loss": 5.0812,
      "mean_token_accuracy": 0.21406900137662888,
      "num_tokens": 19825721.0,
      "step": 8645
    },
    {
      "entropy": 5.278474283218384,
      "epoch": 0.8309317963496637,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004939149915347573,
      "loss": 5.0175,
      "mean_token_accuracy": 0.21694490760564805,
      "num_tokens": 19837273.0,
      "step": 8650
    },
    {
      "entropy": 5.326376581192017,
      "epoch": 0.8314121037463977,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939070697474585,
      "loss": 5.0566,
      "mean_token_accuracy": 0.21395067423582076,
      "num_tokens": 19849436.0,
      "step": 8655
    },
    {
      "entropy": 5.260646724700928,
      "epoch": 0.8318924111431316,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004938991428777875,
      "loss": 4.9332,
      "mean_token_accuracy": 0.2242531105875969,
      "num_tokens": 19860222.0,
      "step": 8660
    },
    {
      "entropy": 5.275563669204712,
      "epoch": 0.8323727185398655,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004938912109259284,
      "loss": 5.0248,
      "mean_token_accuracy": 0.22105076760053635,
      "num_tokens": 19870934.0,
      "step": 8665
    },
    {
      "entropy": 5.329261589050293,
      "epoch": 0.8328530259365994,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004938832738920654,
      "loss": 5.0964,
      "mean_token_accuracy": 0.20505854189395906,
      "num_tokens": 19882149.0,
      "step": 8670
    },
    {
      "entropy": 5.339509582519531,
      "epoch": 0.8333333333333334,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004938753317763826,
      "loss": 5.0247,
      "mean_token_accuracy": 0.21548571884632112,
      "num_tokens": 19893276.0,
      "step": 8675
    },
    {
      "entropy": 5.376326227188111,
      "epoch": 0.8338136407300673,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004938673845790646,
      "loss": 5.1924,
      "mean_token_accuracy": 0.20365149825811385,
      "num_tokens": 19904723.0,
      "step": 8680
    },
    {
      "entropy": 5.3367047786712645,
      "epoch": 0.8342939481268011,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004938594323002957,
      "loss": 5.0916,
      "mean_token_accuracy": 0.20873169898986815,
      "num_tokens": 19915226.0,
      "step": 8685
    },
    {
      "entropy": 5.293796873092651,
      "epoch": 0.834774255523535,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004938514749402609,
      "loss": 5.0846,
      "mean_token_accuracy": 0.21451802402734757,
      "num_tokens": 19927911.0,
      "step": 8690
    },
    {
      "entropy": 5.337069272994995,
      "epoch": 0.835254562920269,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004938435124991447,
      "loss": 5.1535,
      "mean_token_accuracy": 0.2098308712244034,
      "num_tokens": 19940356.0,
      "step": 8695
    },
    {
      "entropy": 5.380351543426514,
      "epoch": 0.8357348703170029,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004938355449771318,
      "loss": 5.1091,
      "mean_token_accuracy": 0.2041558012366295,
      "num_tokens": 19951108.0,
      "step": 8700
    },
    {
      "entropy": 5.29274582862854,
      "epoch": 0.8362151777137368,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004938275723744075,
      "loss": 5.0541,
      "mean_token_accuracy": 0.21939096450805665,
      "num_tokens": 19962427.0,
      "step": 8705
    },
    {
      "entropy": 5.374296522140503,
      "epoch": 0.8366954851104706,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004938195946911567,
      "loss": 5.1592,
      "mean_token_accuracy": 0.20630455315113067,
      "num_tokens": 19973476.0,
      "step": 8710
    },
    {
      "entropy": 5.332964086532593,
      "epoch": 0.8371757925072046,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004938116119275645,
      "loss": 5.0928,
      "mean_token_accuracy": 0.21300906091928482,
      "num_tokens": 19984034.0,
      "step": 8715
    },
    {
      "entropy": 5.386683750152588,
      "epoch": 0.8376560999039385,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004938036240838166,
      "loss": 5.1292,
      "mean_token_accuracy": 0.20694381594657899,
      "num_tokens": 19996035.0,
      "step": 8720
    },
    {
      "entropy": 5.364655685424805,
      "epoch": 0.8381364073006724,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004937956311600983,
      "loss": 5.0532,
      "mean_token_accuracy": 0.21384913325309754,
      "num_tokens": 20007259.0,
      "step": 8725
    },
    {
      "entropy": 5.318237018585205,
      "epoch": 0.8386167146974063,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004937876331565951,
      "loss": 5.0586,
      "mean_token_accuracy": 0.21067376732826232,
      "num_tokens": 20018673.0,
      "step": 8730
    },
    {
      "entropy": 5.285577630996704,
      "epoch": 0.8390970220941403,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004937796300734926,
      "loss": 5.1019,
      "mean_token_accuracy": 0.21162789016962053,
      "num_tokens": 20030767.0,
      "step": 8735
    },
    {
      "entropy": 5.403067922592163,
      "epoch": 0.8395773294908742,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004937716219109769,
      "loss": 5.1277,
      "mean_token_accuracy": 0.20807579755783082,
      "num_tokens": 20041268.0,
      "step": 8740
    },
    {
      "entropy": 5.26431884765625,
      "epoch": 0.840057636887608,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004937636086692336,
      "loss": 4.9854,
      "mean_token_accuracy": 0.22124866098165513,
      "num_tokens": 20052285.0,
      "step": 8745
    },
    {
      "entropy": 5.355442523956299,
      "epoch": 0.840537944284342,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000493755590348449,
      "loss": 5.1974,
      "mean_token_accuracy": 0.20874475836753845,
      "num_tokens": 20065113.0,
      "step": 8750
    },
    {
      "entropy": 5.453869485855103,
      "epoch": 0.8410182516810759,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004937475669488091,
      "loss": 5.1634,
      "mean_token_accuracy": 0.21602114289999008,
      "num_tokens": 20076151.0,
      "step": 8755
    },
    {
      "entropy": 5.323920965194702,
      "epoch": 0.8414985590778098,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004937395384705004,
      "loss": 5.1071,
      "mean_token_accuracy": 0.20864202380180358,
      "num_tokens": 20088195.0,
      "step": 8760
    },
    {
      "entropy": 5.249390983581543,
      "epoch": 0.8419788664745437,
      "grad_norm": 0.875,
      "learning_rate": 0.0004937315049137089,
      "loss": 4.9609,
      "mean_token_accuracy": 0.22439692169427872,
      "num_tokens": 20098576.0,
      "step": 8765
    },
    {
      "entropy": 5.391473865509033,
      "epoch": 0.8424591738712777,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004937234662786216,
      "loss": 5.1745,
      "mean_token_accuracy": 0.20241868495941162,
      "num_tokens": 20110176.0,
      "step": 8770
    },
    {
      "entropy": 5.384263038635254,
      "epoch": 0.8429394812680115,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004937154225654246,
      "loss": 5.0863,
      "mean_token_accuracy": 0.2166321888566017,
      "num_tokens": 20121713.0,
      "step": 8775
    },
    {
      "entropy": 5.432090187072754,
      "epoch": 0.8434197886647454,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004937073737743051,
      "loss": 5.1804,
      "mean_token_accuracy": 0.20397210568189622,
      "num_tokens": 20134126.0,
      "step": 8780
    },
    {
      "entropy": 5.319205665588379,
      "epoch": 0.8439000960614793,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004936993199054499,
      "loss": 5.0671,
      "mean_token_accuracy": 0.21656786650419235,
      "num_tokens": 20144791.0,
      "step": 8785
    },
    {
      "entropy": 5.3982579708099365,
      "epoch": 0.8443804034582133,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004936912609590458,
      "loss": 5.1974,
      "mean_token_accuracy": 0.21172568947076797,
      "num_tokens": 20157214.0,
      "step": 8790
    },
    {
      "entropy": 5.425410604476928,
      "epoch": 0.8448607108549472,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049368319693528,
      "loss": 5.2161,
      "mean_token_accuracy": 0.1998043343424797,
      "num_tokens": 20168952.0,
      "step": 8795
    },
    {
      "entropy": 5.322726249694824,
      "epoch": 0.8453410182516811,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004936751278343397,
      "loss": 5.0897,
      "mean_token_accuracy": 0.20824098438024521,
      "num_tokens": 20181829.0,
      "step": 8800
    },
    {
      "entropy": 5.406108760833741,
      "epoch": 0.845821325648415,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004936670536564123,
      "loss": 5.2319,
      "mean_token_accuracy": 0.1956600695848465,
      "num_tokens": 20193362.0,
      "step": 8805
    },
    {
      "entropy": 5.355117559432983,
      "epoch": 0.8463016330451489,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004936589744016853,
      "loss": 5.1506,
      "mean_token_accuracy": 0.21386872082948685,
      "num_tokens": 20205159.0,
      "step": 8810
    },
    {
      "entropy": 5.310122108459472,
      "epoch": 0.8467819404418828,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000493650890070346,
      "loss": 5.0379,
      "mean_token_accuracy": 0.22063922882080078,
      "num_tokens": 20217075.0,
      "step": 8815
    },
    {
      "entropy": 5.320584440231324,
      "epoch": 0.8472622478386167,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004936428006625824,
      "loss": 5.0461,
      "mean_token_accuracy": 0.220267853140831,
      "num_tokens": 20227901.0,
      "step": 8820
    },
    {
      "entropy": 5.341638803482056,
      "epoch": 0.8477425552353506,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004936347061785823,
      "loss": 5.058,
      "mean_token_accuracy": 0.21537936180830003,
      "num_tokens": 20239945.0,
      "step": 8825
    },
    {
      "entropy": 5.384005165100097,
      "epoch": 0.8482228626320846,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004936266066185334,
      "loss": 5.1516,
      "mean_token_accuracy": 0.20924608707427977,
      "num_tokens": 20251503.0,
      "step": 8830
    },
    {
      "entropy": 5.32397346496582,
      "epoch": 0.8487031700288185,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004936185019826239,
      "loss": 5.0583,
      "mean_token_accuracy": 0.21213850378990173,
      "num_tokens": 20263044.0,
      "step": 8835
    },
    {
      "entropy": 5.345823955535889,
      "epoch": 0.8491834774255523,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004936103922710419,
      "loss": 5.0415,
      "mean_token_accuracy": 0.21659981310367585,
      "num_tokens": 20273563.0,
      "step": 8840
    },
    {
      "entropy": 5.33765172958374,
      "epoch": 0.8496637848222862,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004936022774839759,
      "loss": 5.0958,
      "mean_token_accuracy": 0.20943790525197983,
      "num_tokens": 20285482.0,
      "step": 8845
    },
    {
      "entropy": 5.435449314117432,
      "epoch": 0.8501440922190202,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004935941576216141,
      "loss": 5.281,
      "mean_token_accuracy": 0.20362711399793626,
      "num_tokens": 20296666.0,
      "step": 8850
    },
    {
      "entropy": 5.236143255233765,
      "epoch": 0.8506243996157541,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000493586032684145,
      "loss": 4.9581,
      "mean_token_accuracy": 0.22281866073608397,
      "num_tokens": 20308978.0,
      "step": 8855
    },
    {
      "entropy": 5.296805047988892,
      "epoch": 0.851104707012488,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004935779026717573,
      "loss": 5.032,
      "mean_token_accuracy": 0.21854369938373566,
      "num_tokens": 20321488.0,
      "step": 8860
    },
    {
      "entropy": 5.267953252792358,
      "epoch": 0.8515850144092219,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004935697675846396,
      "loss": 4.8993,
      "mean_token_accuracy": 0.2284442588686943,
      "num_tokens": 20332938.0,
      "step": 8865
    },
    {
      "entropy": 5.256423902511597,
      "epoch": 0.8520653218059558,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004935616274229811,
      "loss": 5.0163,
      "mean_token_accuracy": 0.22189441621303557,
      "num_tokens": 20342986.0,
      "step": 8870
    },
    {
      "entropy": 5.304368925094605,
      "epoch": 0.8525456292026897,
      "grad_norm": 0.875,
      "learning_rate": 0.0004935534821869705,
      "loss": 5.044,
      "mean_token_accuracy": 0.21364661902189255,
      "num_tokens": 20355791.0,
      "step": 8875
    },
    {
      "entropy": 5.238782215118408,
      "epoch": 0.8530259365994236,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004935453318767971,
      "loss": 4.939,
      "mean_token_accuracy": 0.21945467591285706,
      "num_tokens": 20367080.0,
      "step": 8880
    },
    {
      "entropy": 5.332766103744507,
      "epoch": 0.8535062439961575,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00049353717649265,
      "loss": 5.0396,
      "mean_token_accuracy": 0.223314169049263,
      "num_tokens": 20376697.0,
      "step": 8885
    },
    {
      "entropy": 5.338983488082886,
      "epoch": 0.8539865513928915,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004935290160347185,
      "loss": 5.0718,
      "mean_token_accuracy": 0.21374462693929672,
      "num_tokens": 20387430.0,
      "step": 8890
    },
    {
      "entropy": 5.294669961929321,
      "epoch": 0.8544668587896254,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004935208505031922,
      "loss": 5.042,
      "mean_token_accuracy": 0.2144953116774559,
      "num_tokens": 20398488.0,
      "step": 8895
    },
    {
      "entropy": 5.236431741714478,
      "epoch": 0.8549471661863592,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004935126798982606,
      "loss": 5.0848,
      "mean_token_accuracy": 0.21171467900276184,
      "num_tokens": 20410316.0,
      "step": 8900
    },
    {
      "entropy": 5.341539716720581,
      "epoch": 0.8554274735830932,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004935045042201135,
      "loss": 4.9963,
      "mean_token_accuracy": 0.22462246865034102,
      "num_tokens": 20421645.0,
      "step": 8905
    },
    {
      "entropy": 5.363277387619019,
      "epoch": 0.8559077809798271,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004934963234689407,
      "loss": 5.1108,
      "mean_token_accuracy": 0.20586680471897126,
      "num_tokens": 20433397.0,
      "step": 8910
    },
    {
      "entropy": 5.246917295455932,
      "epoch": 0.856388088376561,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000493488137644932,
      "loss": 5.0231,
      "mean_token_accuracy": 0.216941174864769,
      "num_tokens": 20446218.0,
      "step": 8915
    },
    {
      "entropy": 5.363285398483276,
      "epoch": 0.8568683957732949,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004934799467482774,
      "loss": 5.1035,
      "mean_token_accuracy": 0.2093177244067192,
      "num_tokens": 20457265.0,
      "step": 8920
    },
    {
      "entropy": 5.309921360015869,
      "epoch": 0.8573487031700289,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004934717507791673,
      "loss": 5.0966,
      "mean_token_accuracy": 0.21139997094869614,
      "num_tokens": 20468748.0,
      "step": 8925
    },
    {
      "entropy": 5.33907470703125,
      "epoch": 0.8578290105667628,
      "grad_norm": 0.875,
      "learning_rate": 0.0004934635497377919,
      "loss": 5.1011,
      "mean_token_accuracy": 0.21302476078271865,
      "num_tokens": 20481416.0,
      "step": 8930
    },
    {
      "entropy": 5.250659561157226,
      "epoch": 0.8583093179634966,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004934553436243415,
      "loss": 5.0139,
      "mean_token_accuracy": 0.21668749898672104,
      "num_tokens": 20493063.0,
      "step": 8935
    },
    {
      "entropy": 5.277139949798584,
      "epoch": 0.8587896253602305,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004934471324390067,
      "loss": 5.0029,
      "mean_token_accuracy": 0.214878810942173,
      "num_tokens": 20504000.0,
      "step": 8940
    },
    {
      "entropy": 5.349289751052856,
      "epoch": 0.8592699327569645,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004934389161819783,
      "loss": 5.0996,
      "mean_token_accuracy": 0.21724836826324462,
      "num_tokens": 20516419.0,
      "step": 8945
    },
    {
      "entropy": 5.321029758453369,
      "epoch": 0.8597502401536984,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004934306948534467,
      "loss": 5.0641,
      "mean_token_accuracy": 0.21116738319396972,
      "num_tokens": 20527385.0,
      "step": 8950
    },
    {
      "entropy": 5.305115842819214,
      "epoch": 0.8602305475504323,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004934224684536031,
      "loss": 5.0513,
      "mean_token_accuracy": 0.21651874333620072,
      "num_tokens": 20538051.0,
      "step": 8955
    },
    {
      "entropy": 5.347703123092652,
      "epoch": 0.8607108549471661,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004934142369826382,
      "loss": 5.0672,
      "mean_token_accuracy": 0.21298073083162308,
      "num_tokens": 20550321.0,
      "step": 8960
    },
    {
      "entropy": 5.340742206573486,
      "epoch": 0.8611911623439001,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004934060004407434,
      "loss": 5.0636,
      "mean_token_accuracy": 0.21349679231643676,
      "num_tokens": 20561229.0,
      "step": 8965
    },
    {
      "entropy": 5.272282552719116,
      "epoch": 0.861671469740634,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004933977588281099,
      "loss": 5.0175,
      "mean_token_accuracy": 0.21788799166679382,
      "num_tokens": 20572040.0,
      "step": 8970
    },
    {
      "entropy": 5.251030206680298,
      "epoch": 0.8621517771373679,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004933895121449288,
      "loss": 5.047,
      "mean_token_accuracy": 0.21450204402208328,
      "num_tokens": 20583110.0,
      "step": 8975
    },
    {
      "entropy": 5.279579162597656,
      "epoch": 0.8626320845341018,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004933812603913917,
      "loss": 5.0341,
      "mean_token_accuracy": 0.21116321682929992,
      "num_tokens": 20593647.0,
      "step": 8980
    },
    {
      "entropy": 5.244159603118897,
      "epoch": 0.8631123919308358,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004933730035676903,
      "loss": 4.9163,
      "mean_token_accuracy": 0.2284306988120079,
      "num_tokens": 20604428.0,
      "step": 8985
    },
    {
      "entropy": 5.2218774318695065,
      "epoch": 0.8635926993275697,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004933647416740161,
      "loss": 4.9798,
      "mean_token_accuracy": 0.2260243207216263,
      "num_tokens": 20615811.0,
      "step": 8990
    },
    {
      "entropy": 5.372859811782837,
      "epoch": 0.8640730067243035,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000493356474710561,
      "loss": 5.0739,
      "mean_token_accuracy": 0.21329560577869416,
      "num_tokens": 20627679.0,
      "step": 8995
    },
    {
      "entropy": 5.381857204437256,
      "epoch": 0.8645533141210374,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000493348202677517,
      "loss": 5.1045,
      "mean_token_accuracy": 0.21503155529499055,
      "num_tokens": 20638879.0,
      "step": 9000
    },
    {
      "epoch": 0.8645533141210374,
      "eval_entropy": 5.127662898272436,
      "eval_loss": 5.085931301116943,
      "eval_mean_token_accuracy": 0.22127145906408086,
      "eval_num_tokens": 20638879.0,
      "eval_runtime": 20.7356,
      "eval_samples_per_second": 1582.544,
      "eval_steps_per_second": 197.824,
      "step": 9000
    },
    {
      "entropy": 5.311496114730835,
      "epoch": 0.8650336215177714,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004933399255750761,
      "loss": 5.0906,
      "mean_token_accuracy": 0.22049616128206254,
      "num_tokens": 20649729.0,
      "step": 9005
    },
    {
      "entropy": 5.3861250400543215,
      "epoch": 0.8655139289145053,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004933316434034304,
      "loss": 5.1399,
      "mean_token_accuracy": 0.22092850804328917,
      "num_tokens": 20660473.0,
      "step": 9010
    },
    {
      "entropy": 5.362330484390259,
      "epoch": 0.8659942363112392,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004933233561627723,
      "loss": 5.108,
      "mean_token_accuracy": 0.2110399916768074,
      "num_tokens": 20671776.0,
      "step": 9015
    },
    {
      "entropy": 5.28601040840149,
      "epoch": 0.866474543707973,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004933150638532942,
      "loss": 5.0802,
      "mean_token_accuracy": 0.21654984951019288,
      "num_tokens": 20684147.0,
      "step": 9020
    },
    {
      "entropy": 5.300482273101807,
      "epoch": 0.866954851104707,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004933067664751885,
      "loss": 5.0286,
      "mean_token_accuracy": 0.2167188674211502,
      "num_tokens": 20695248.0,
      "step": 9025
    },
    {
      "entropy": 5.311449003219605,
      "epoch": 0.8674351585014409,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000493298464028648,
      "loss": 5.0635,
      "mean_token_accuracy": 0.21684966683387757,
      "num_tokens": 20707772.0,
      "step": 9030
    },
    {
      "entropy": 5.341923189163208,
      "epoch": 0.8679154658981748,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004932901565138653,
      "loss": 5.0338,
      "mean_token_accuracy": 0.2088928446173668,
      "num_tokens": 20718813.0,
      "step": 9035
    },
    {
      "entropy": 5.27556266784668,
      "epoch": 0.8683957732949087,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004932818439310334,
      "loss": 5.0196,
      "mean_token_accuracy": 0.22322826981544494,
      "num_tokens": 20730939.0,
      "step": 9040
    },
    {
      "entropy": 5.384440469741821,
      "epoch": 0.8688760806916427,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004932735262803452,
      "loss": 5.1224,
      "mean_token_accuracy": 0.2094522088766098,
      "num_tokens": 20742990.0,
      "step": 9045
    },
    {
      "entropy": 5.271628952026367,
      "epoch": 0.8693563880883766,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004932652035619939,
      "loss": 4.9868,
      "mean_token_accuracy": 0.21627162843942643,
      "num_tokens": 20754076.0,
      "step": 9050
    },
    {
      "entropy": 5.299597215652466,
      "epoch": 0.8698366954851104,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004932568757761727,
      "loss": 5.0264,
      "mean_token_accuracy": 0.2147304058074951,
      "num_tokens": 20765538.0,
      "step": 9055
    },
    {
      "entropy": 5.335405921936035,
      "epoch": 0.8703170028818443,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004932485429230748,
      "loss": 5.1036,
      "mean_token_accuracy": 0.20875843316316606,
      "num_tokens": 20776359.0,
      "step": 9060
    },
    {
      "entropy": 5.28267273902893,
      "epoch": 0.8707973102785783,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000493240205002894,
      "loss": 5.0839,
      "mean_token_accuracy": 0.21629297882318496,
      "num_tokens": 20787581.0,
      "step": 9065
    },
    {
      "entropy": 5.388483238220215,
      "epoch": 0.8712776176753122,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004932318620158235,
      "loss": 5.1145,
      "mean_token_accuracy": 0.21012043356895446,
      "num_tokens": 20799904.0,
      "step": 9070
    },
    {
      "entropy": 5.360992383956909,
      "epoch": 0.8717579250720461,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004932235139620574,
      "loss": 5.0409,
      "mean_token_accuracy": 0.2205362558364868,
      "num_tokens": 20810238.0,
      "step": 9075
    },
    {
      "entropy": 5.276711320877075,
      "epoch": 0.8722382324687801,
      "grad_norm": 1.0,
      "learning_rate": 0.0004932151608417892,
      "loss": 4.9852,
      "mean_token_accuracy": 0.21606039851903916,
      "num_tokens": 20821349.0,
      "step": 9080
    },
    {
      "entropy": 5.2188639640808105,
      "epoch": 0.872718539865514,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004932068026552127,
      "loss": 5.0343,
      "mean_token_accuracy": 0.2166193261742592,
      "num_tokens": 20834788.0,
      "step": 9085
    },
    {
      "entropy": 5.3543306350708,
      "epoch": 0.8731988472622478,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004931984394025224,
      "loss": 5.1086,
      "mean_token_accuracy": 0.2120328053832054,
      "num_tokens": 20845571.0,
      "step": 9090
    },
    {
      "entropy": 5.398407602310181,
      "epoch": 0.8736791546589817,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004931900710839123,
      "loss": 5.0985,
      "mean_token_accuracy": 0.2099953070282936,
      "num_tokens": 20857209.0,
      "step": 9095
    },
    {
      "entropy": 5.389593982696534,
      "epoch": 0.8741594620557157,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004931816976995766,
      "loss": 5.1507,
      "mean_token_accuracy": 0.21171553134918214,
      "num_tokens": 20870624.0,
      "step": 9100
    },
    {
      "entropy": 5.280370664596558,
      "epoch": 0.8746397694524496,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004931733192497097,
      "loss": 5.059,
      "mean_token_accuracy": 0.21613684743642808,
      "num_tokens": 20881769.0,
      "step": 9105
    },
    {
      "entropy": 5.25521183013916,
      "epoch": 0.8751200768491835,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004931649357345062,
      "loss": 5.0224,
      "mean_token_accuracy": 0.21273586750030518,
      "num_tokens": 20892817.0,
      "step": 9110
    },
    {
      "entropy": 5.328509855270386,
      "epoch": 0.8756003842459174,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004931565471541606,
      "loss": 4.9993,
      "mean_token_accuracy": 0.21624964475631714,
      "num_tokens": 20903042.0,
      "step": 9115
    },
    {
      "entropy": 5.23095006942749,
      "epoch": 0.8760806916426513,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004931481535088679,
      "loss": 4.9499,
      "mean_token_accuracy": 0.22801749557256698,
      "num_tokens": 20914684.0,
      "step": 9120
    },
    {
      "entropy": 5.237929964065552,
      "epoch": 0.8765609990393852,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004931397547988229,
      "loss": 5.0191,
      "mean_token_accuracy": 0.22458050400018692,
      "num_tokens": 20926585.0,
      "step": 9125
    },
    {
      "entropy": 5.356379175186158,
      "epoch": 0.8770413064361191,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004931313510242204,
      "loss": 5.0704,
      "mean_token_accuracy": 0.21569012105464935,
      "num_tokens": 20939729.0,
      "step": 9130
    },
    {
      "entropy": 5.27813024520874,
      "epoch": 0.877521613832853,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004931229421852557,
      "loss": 4.9982,
      "mean_token_accuracy": 0.217333023250103,
      "num_tokens": 20951697.0,
      "step": 9135
    },
    {
      "entropy": 5.3300739288330075,
      "epoch": 0.878001921229587,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000493114528282124,
      "loss": 5.1346,
      "mean_token_accuracy": 0.210756878554821,
      "num_tokens": 20962729.0,
      "step": 9140
    },
    {
      "entropy": 5.2643290042877195,
      "epoch": 0.8784822286263209,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004931061093150206,
      "loss": 4.9948,
      "mean_token_accuracy": 0.22450024634599686,
      "num_tokens": 20973331.0,
      "step": 9145
    },
    {
      "entropy": 5.24806776046753,
      "epoch": 0.8789625360230547,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004930976852841409,
      "loss": 4.9969,
      "mean_token_accuracy": 0.20946622192859649,
      "num_tokens": 20985609.0,
      "step": 9150
    },
    {
      "entropy": 5.2708662986755375,
      "epoch": 0.8794428434197886,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004930892561896806,
      "loss": 5.0174,
      "mean_token_accuracy": 0.21318918615579605,
      "num_tokens": 20997231.0,
      "step": 9155
    },
    {
      "entropy": 5.323353052139282,
      "epoch": 0.8799231508165226,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004930808220318354,
      "loss": 5.0424,
      "mean_token_accuracy": 0.21940839290618896,
      "num_tokens": 21008511.0,
      "step": 9160
    },
    {
      "entropy": 5.258856916427613,
      "epoch": 0.8804034582132565,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004930723828108012,
      "loss": 4.981,
      "mean_token_accuracy": 0.2166222780942917,
      "num_tokens": 21019108.0,
      "step": 9165
    },
    {
      "entropy": 5.278417348861694,
      "epoch": 0.8808837656099904,
      "grad_norm": 1.0,
      "learning_rate": 0.0004930639385267736,
      "loss": 5.0459,
      "mean_token_accuracy": 0.2172716036438942,
      "num_tokens": 21030621.0,
      "step": 9170
    },
    {
      "entropy": 5.286413621902466,
      "epoch": 0.8813640730067243,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000493055489179949,
      "loss": 5.0044,
      "mean_token_accuracy": 0.217023766040802,
      "num_tokens": 21041778.0,
      "step": 9175
    },
    {
      "entropy": 5.304246711730957,
      "epoch": 0.8818443804034583,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004930470347705234,
      "loss": 5.0199,
      "mean_token_accuracy": 0.22323419749736786,
      "num_tokens": 21054257.0,
      "step": 9180
    },
    {
      "entropy": 5.326455307006836,
      "epoch": 0.8823246878001921,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000493038575298693,
      "loss": 5.0756,
      "mean_token_accuracy": 0.20766589492559434,
      "num_tokens": 21066378.0,
      "step": 9185
    },
    {
      "entropy": 5.23599100112915,
      "epoch": 0.882804995196926,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004930301107646545,
      "loss": 5.0016,
      "mean_token_accuracy": 0.21614992916584014,
      "num_tokens": 21078913.0,
      "step": 9190
    },
    {
      "entropy": 5.335093784332275,
      "epoch": 0.8832853025936599,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004930216411686042,
      "loss": 5.0663,
      "mean_token_accuracy": 0.21228698790073394,
      "num_tokens": 21090500.0,
      "step": 9195
    },
    {
      "entropy": 5.27804970741272,
      "epoch": 0.8837656099903939,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004930131665107387,
      "loss": 5.0171,
      "mean_token_accuracy": 0.2161962553858757,
      "num_tokens": 21102793.0,
      "step": 9200
    },
    {
      "entropy": 5.3840419292449955,
      "epoch": 0.8842459173871278,
      "grad_norm": 0.953125,
      "learning_rate": 0.000493004686791255,
      "loss": 5.1312,
      "mean_token_accuracy": 0.20459095686674117,
      "num_tokens": 21114504.0,
      "step": 9205
    },
    {
      "entropy": 5.299411678314209,
      "epoch": 0.8847262247838616,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004929962020103496,
      "loss": 5.006,
      "mean_token_accuracy": 0.214769646525383,
      "num_tokens": 21126733.0,
      "step": 9210
    },
    {
      "entropy": 5.260265254974366,
      "epoch": 0.8852065321805955,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004929877121682198,
      "loss": 5.0062,
      "mean_token_accuracy": 0.2167923390865326,
      "num_tokens": 21138045.0,
      "step": 9215
    },
    {
      "entropy": 5.2883553981781,
      "epoch": 0.8856868395773295,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004929792172650627,
      "loss": 5.0821,
      "mean_token_accuracy": 0.2079818457365036,
      "num_tokens": 21151562.0,
      "step": 9220
    },
    {
      "entropy": 5.317444467544556,
      "epoch": 0.8861671469740634,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004929707173010753,
      "loss": 5.0488,
      "mean_token_accuracy": 0.2134185954928398,
      "num_tokens": 21162943.0,
      "step": 9225
    },
    {
      "entropy": 5.33593225479126,
      "epoch": 0.8866474543707973,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004929622122764552,
      "loss": 5.0783,
      "mean_token_accuracy": 0.21358013451099395,
      "num_tokens": 21174392.0,
      "step": 9230
    },
    {
      "entropy": 5.253098869323731,
      "epoch": 0.8871277617675313,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004929537021913997,
      "loss": 4.9193,
      "mean_token_accuracy": 0.22413647323846816,
      "num_tokens": 21185372.0,
      "step": 9235
    },
    {
      "entropy": 5.303736925125122,
      "epoch": 0.8876080691642652,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004929451870461064,
      "loss": 5.0677,
      "mean_token_accuracy": 0.21268622130155562,
      "num_tokens": 21197044.0,
      "step": 9240
    },
    {
      "entropy": 5.253814649581909,
      "epoch": 0.888088376560999,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004929366668407731,
      "loss": 5.0066,
      "mean_token_accuracy": 0.21816504448652269,
      "num_tokens": 21207729.0,
      "step": 9245
    },
    {
      "entropy": 5.344317817687989,
      "epoch": 0.8885686839577329,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004929281415755974,
      "loss": 5.0398,
      "mean_token_accuracy": 0.21348556280136108,
      "num_tokens": 21218909.0,
      "step": 9250
    },
    {
      "entropy": 5.366525888442993,
      "epoch": 0.8890489913544669,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004929196112507775,
      "loss": 5.1501,
      "mean_token_accuracy": 0.2105141580104828,
      "num_tokens": 21230543.0,
      "step": 9255
    },
    {
      "entropy": 5.321249961853027,
      "epoch": 0.8895292987512008,
      "grad_norm": 1.0,
      "learning_rate": 0.0004929110758665112,
      "loss": 5.0841,
      "mean_token_accuracy": 0.21225471496582032,
      "num_tokens": 21242064.0,
      "step": 9260
    },
    {
      "entropy": 5.356348657608033,
      "epoch": 0.8900096061479347,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004929025354229969,
      "loss": 5.112,
      "mean_token_accuracy": 0.208681683242321,
      "num_tokens": 21254321.0,
      "step": 9265
    },
    {
      "entropy": 5.23546290397644,
      "epoch": 0.8904899135446686,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004928939899204326,
      "loss": 4.9357,
      "mean_token_accuracy": 0.22330885082483293,
      "num_tokens": 21264741.0,
      "step": 9270
    },
    {
      "entropy": 5.308848714828491,
      "epoch": 0.8909702209414025,
      "grad_norm": 0.921875,
      "learning_rate": 0.000492885439359017,
      "loss": 5.0299,
      "mean_token_accuracy": 0.21464652568101883,
      "num_tokens": 21276834.0,
      "step": 9275
    },
    {
      "entropy": 5.237514019012451,
      "epoch": 0.8914505283381364,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004928768837389485,
      "loss": 5.0116,
      "mean_token_accuracy": 0.21696730703115463,
      "num_tokens": 21287108.0,
      "step": 9280
    },
    {
      "entropy": 5.264032649993896,
      "epoch": 0.8919308357348703,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004928683230604257,
      "loss": 5.0152,
      "mean_token_accuracy": 0.21518100798130035,
      "num_tokens": 21299942.0,
      "step": 9285
    },
    {
      "entropy": 5.428152656555175,
      "epoch": 0.8924111431316042,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004928597573236474,
      "loss": 5.1643,
      "mean_token_accuracy": 0.20927155315876006,
      "num_tokens": 21311243.0,
      "step": 9290
    },
    {
      "entropy": 5.329810094833374,
      "epoch": 0.8928914505283382,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004928511865288123,
      "loss": 4.9914,
      "mean_token_accuracy": 0.2204737886786461,
      "num_tokens": 21322291.0,
      "step": 9295
    },
    {
      "entropy": 5.284492778778076,
      "epoch": 0.8933717579250721,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004928426106761197,
      "loss": 5.1114,
      "mean_token_accuracy": 0.21199363321065903,
      "num_tokens": 21333257.0,
      "step": 9300
    },
    {
      "entropy": 5.230852937698364,
      "epoch": 0.8938520653218059,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004928340297657685,
      "loss": 4.9979,
      "mean_token_accuracy": 0.2132027953863144,
      "num_tokens": 21345848.0,
      "step": 9305
    },
    {
      "entropy": 5.334110260009766,
      "epoch": 0.8943323727185398,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004928254437979578,
      "loss": 5.0301,
      "mean_token_accuracy": 0.22159041315317154,
      "num_tokens": 21357693.0,
      "step": 9310
    },
    {
      "entropy": 5.321630239486694,
      "epoch": 0.8948126801152738,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004928168527728873,
      "loss": 5.1434,
      "mean_token_accuracy": 0.20822151899337768,
      "num_tokens": 21369653.0,
      "step": 9315
    },
    {
      "entropy": 5.38863615989685,
      "epoch": 0.8952929875120077,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004928082566907562,
      "loss": 5.0742,
      "mean_token_accuracy": 0.21584334075450898,
      "num_tokens": 21383924.0,
      "step": 9320
    },
    {
      "entropy": 5.340769815444946,
      "epoch": 0.8957732949087416,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004927996555517642,
      "loss": 5.0604,
      "mean_token_accuracy": 0.21026284992694855,
      "num_tokens": 21395963.0,
      "step": 9325
    },
    {
      "entropy": 5.299890947341919,
      "epoch": 0.8962536023054755,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004927910493561109,
      "loss": 5.0561,
      "mean_token_accuracy": 0.21581374406814574,
      "num_tokens": 21408200.0,
      "step": 9330
    },
    {
      "entropy": 5.369954538345337,
      "epoch": 0.8967339097022095,
      "grad_norm": 0.9375,
      "learning_rate": 0.000492782438103996,
      "loss": 5.1714,
      "mean_token_accuracy": 0.2066921442747116,
      "num_tokens": 21419963.0,
      "step": 9335
    },
    {
      "entropy": 5.348271608352661,
      "epoch": 0.8972142170989433,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004927738217956197,
      "loss": 5.1222,
      "mean_token_accuracy": 0.2084527373313904,
      "num_tokens": 21431824.0,
      "step": 9340
    },
    {
      "entropy": 5.211304807662964,
      "epoch": 0.8976945244956772,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004927652004311819,
      "loss": 4.8885,
      "mean_token_accuracy": 0.22600077390670775,
      "num_tokens": 21442354.0,
      "step": 9345
    },
    {
      "entropy": 5.340343236923218,
      "epoch": 0.8981748318924111,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004927565740108828,
      "loss": 5.0451,
      "mean_token_accuracy": 0.21112367361783982,
      "num_tokens": 21453734.0,
      "step": 9350
    },
    {
      "entropy": 5.326572275161743,
      "epoch": 0.8986551392891451,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004927479425349226,
      "loss": 5.0582,
      "mean_token_accuracy": 0.21704673320055007,
      "num_tokens": 21465471.0,
      "step": 9355
    },
    {
      "entropy": 5.361413621902466,
      "epoch": 0.899135446685879,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004927393060035018,
      "loss": 5.1929,
      "mean_token_accuracy": 0.20296454727649688,
      "num_tokens": 21477775.0,
      "step": 9360
    },
    {
      "entropy": 5.242503976821899,
      "epoch": 0.8996157540826129,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004927306644168207,
      "loss": 4.9659,
      "mean_token_accuracy": 0.22754594087600707,
      "num_tokens": 21489319.0,
      "step": 9365
    },
    {
      "entropy": 5.417535495758057,
      "epoch": 0.9000960614793467,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004927220177750803,
      "loss": 5.2106,
      "mean_token_accuracy": 0.20667925029993056,
      "num_tokens": 21499742.0,
      "step": 9370
    },
    {
      "entropy": 5.307334280014038,
      "epoch": 0.9005763688760807,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004927133660784811,
      "loss": 4.983,
      "mean_token_accuracy": 0.2206253856420517,
      "num_tokens": 21511063.0,
      "step": 9375
    },
    {
      "entropy": 5.297605180740357,
      "epoch": 0.9010566762728146,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004927047093272241,
      "loss": 5.0047,
      "mean_token_accuracy": 0.21953643411397933,
      "num_tokens": 21522500.0,
      "step": 9380
    },
    {
      "entropy": 5.338352870941162,
      "epoch": 0.9015369836695485,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00049269604752151,
      "loss": 5.0856,
      "mean_token_accuracy": 0.21047282069921494,
      "num_tokens": 21533578.0,
      "step": 9385
    },
    {
      "entropy": 5.321967506408692,
      "epoch": 0.9020172910662824,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004926873806615403,
      "loss": 5.0759,
      "mean_token_accuracy": 0.20933518558740616,
      "num_tokens": 21544296.0,
      "step": 9390
    },
    {
      "entropy": 5.344237518310547,
      "epoch": 0.9024975984630164,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004926787087475158,
      "loss": 5.1508,
      "mean_token_accuracy": 0.2056669071316719,
      "num_tokens": 21555386.0,
      "step": 9395
    },
    {
      "entropy": 5.285631275177002,
      "epoch": 0.9029779058597502,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004926700317796382,
      "loss": 4.9279,
      "mean_token_accuracy": 0.22984151244163514,
      "num_tokens": 21566527.0,
      "step": 9400
    },
    {
      "entropy": 5.412722492218018,
      "epoch": 0.9034582132564841,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004926613497581088,
      "loss": 5.076,
      "mean_token_accuracy": 0.2087673708796501,
      "num_tokens": 21576870.0,
      "step": 9405
    },
    {
      "entropy": 5.333259010314942,
      "epoch": 0.9039385206532181,
      "grad_norm": 0.875,
      "learning_rate": 0.0004926526626831292,
      "loss": 5.0399,
      "mean_token_accuracy": 0.21628521531820297,
      "num_tokens": 21588113.0,
      "step": 9410
    },
    {
      "entropy": 5.295807027816773,
      "epoch": 0.904418828049952,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004926439705549011,
      "loss": 5.0499,
      "mean_token_accuracy": 0.21196681410074233,
      "num_tokens": 21599307.0,
      "step": 9415
    },
    {
      "entropy": 5.303540945053101,
      "epoch": 0.9048991354466859,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004926352733736262,
      "loss": 5.001,
      "mean_token_accuracy": 0.2217928484082222,
      "num_tokens": 21609961.0,
      "step": 9420
    },
    {
      "entropy": 5.2638544082641605,
      "epoch": 0.9053794428434198,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004926265711395065,
      "loss": 5.0421,
      "mean_token_accuracy": 0.21590722799301149,
      "num_tokens": 21622222.0,
      "step": 9425
    },
    {
      "entropy": 5.298861026763916,
      "epoch": 0.9058597502401537,
      "grad_norm": 0.9375,
      "learning_rate": 0.000492617863852744,
      "loss": 5.0159,
      "mean_token_accuracy": 0.22202509343624116,
      "num_tokens": 21632843.0,
      "step": 9430
    },
    {
      "entropy": 5.322651386260986,
      "epoch": 0.9063400576368876,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004926091515135409,
      "loss": 5.0686,
      "mean_token_accuracy": 0.21114126741886138,
      "num_tokens": 21645387.0,
      "step": 9435
    },
    {
      "entropy": 5.327975845336914,
      "epoch": 0.9068203650336215,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004926004341220995,
      "loss": 5.0306,
      "mean_token_accuracy": 0.21964469701051711,
      "num_tokens": 21656787.0,
      "step": 9440
    },
    {
      "entropy": 5.328755617141724,
      "epoch": 0.9073006724303554,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004925917116786222,
      "loss": 5.0564,
      "mean_token_accuracy": 0.21183840483427047,
      "num_tokens": 21667800.0,
      "step": 9445
    },
    {
      "entropy": 5.349583292007447,
      "epoch": 0.9077809798270894,
      "grad_norm": 1.0,
      "learning_rate": 0.0004925829841833114,
      "loss": 5.1062,
      "mean_token_accuracy": 0.20923036187887192,
      "num_tokens": 21679297.0,
      "step": 9450
    },
    {
      "entropy": 5.347641801834106,
      "epoch": 0.9082612872238233,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004925742516363699,
      "loss": 5.1108,
      "mean_token_accuracy": 0.2061640813946724,
      "num_tokens": 21692956.0,
      "step": 9455
    },
    {
      "entropy": 5.290804290771485,
      "epoch": 0.9087415946205571,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004925655140380002,
      "loss": 4.892,
      "mean_token_accuracy": 0.22556530982255935,
      "num_tokens": 21704852.0,
      "step": 9460
    },
    {
      "entropy": 5.308605909347534,
      "epoch": 0.909221902017291,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004925567713884054,
      "loss": 5.0407,
      "mean_token_accuracy": 0.21473145186901094,
      "num_tokens": 21715886.0,
      "step": 9465
    },
    {
      "entropy": 5.283479070663452,
      "epoch": 0.909702209414025,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004925480236877884,
      "loss": 5.0087,
      "mean_token_accuracy": 0.21911457031965256,
      "num_tokens": 21727157.0,
      "step": 9470
    },
    {
      "entropy": 5.333270740509033,
      "epoch": 0.9101825168107589,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004925392709363522,
      "loss": 5.0377,
      "mean_token_accuracy": 0.21460600048303605,
      "num_tokens": 21738232.0,
      "step": 9475
    },
    {
      "entropy": 5.288477373123169,
      "epoch": 0.9106628242074928,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004925305131343001,
      "loss": 5.0589,
      "mean_token_accuracy": 0.21448163837194442,
      "num_tokens": 21749029.0,
      "step": 9480
    },
    {
      "entropy": 5.3516990661621096,
      "epoch": 0.9111431316042267,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004925217502818355,
      "loss": 5.0744,
      "mean_token_accuracy": 0.21320670396089553,
      "num_tokens": 21761243.0,
      "step": 9485
    },
    {
      "entropy": 5.270552253723144,
      "epoch": 0.9116234390009607,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004925129823791616,
      "loss": 5.0398,
      "mean_token_accuracy": 0.21175860464572907,
      "num_tokens": 21772623.0,
      "step": 9490
    },
    {
      "entropy": 5.358375358581543,
      "epoch": 0.9121037463976945,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004925042094264822,
      "loss": 5.0601,
      "mean_token_accuracy": 0.21803104430437087,
      "num_tokens": 21782611.0,
      "step": 9495
    },
    {
      "entropy": 5.237972259521484,
      "epoch": 0.9125840537944284,
      "grad_norm": 1.015625,
      "learning_rate": 0.000492495431424001,
      "loss": 4.93,
      "mean_token_accuracy": 0.22726993262767792,
      "num_tokens": 21793946.0,
      "step": 9500
    },
    {
      "entropy": 5.393427562713623,
      "epoch": 0.9130643611911623,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004924866483719216,
      "loss": 5.1295,
      "mean_token_accuracy": 0.20352391451597213,
      "num_tokens": 21803878.0,
      "step": 9505
    },
    {
      "entropy": 5.306346750259399,
      "epoch": 0.9135446685878963,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004924778602704481,
      "loss": 5.0548,
      "mean_token_accuracy": 0.21506883203983307,
      "num_tokens": 21815187.0,
      "step": 9510
    },
    {
      "entropy": 5.3032678127288815,
      "epoch": 0.9140249759846302,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004924690671197845,
      "loss": 5.0314,
      "mean_token_accuracy": 0.21780543327331542,
      "num_tokens": 21825597.0,
      "step": 9515
    },
    {
      "entropy": 5.273237991333008,
      "epoch": 0.914505283381364,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004924602689201348,
      "loss": 5.005,
      "mean_token_accuracy": 0.2193635731935501,
      "num_tokens": 21837110.0,
      "step": 9520
    },
    {
      "entropy": 5.494055557250976,
      "epoch": 0.9149855907780979,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004924514656717034,
      "loss": 5.2637,
      "mean_token_accuracy": 0.20616769790649414,
      "num_tokens": 21847754.0,
      "step": 9525
    },
    {
      "entropy": 5.291819763183594,
      "epoch": 0.9154658981748319,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004924426573746948,
      "loss": 5.0491,
      "mean_token_accuracy": 0.21123050451278685,
      "num_tokens": 21859162.0,
      "step": 9530
    },
    {
      "entropy": 5.330890464782715,
      "epoch": 0.9159462055715658,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004924338440293131,
      "loss": 5.0156,
      "mean_token_accuracy": 0.21661546230316162,
      "num_tokens": 21870826.0,
      "step": 9535
    },
    {
      "entropy": 5.3212769508361815,
      "epoch": 0.9164265129682997,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004924250256357635,
      "loss": 5.0664,
      "mean_token_accuracy": 0.2101300537586212,
      "num_tokens": 21882003.0,
      "step": 9540
    },
    {
      "entropy": 5.295588207244873,
      "epoch": 0.9169068203650336,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004924162021942502,
      "loss": 5.0113,
      "mean_token_accuracy": 0.21557917445898056,
      "num_tokens": 21894132.0,
      "step": 9545
    },
    {
      "entropy": 5.276459789276123,
      "epoch": 0.9173871277617676,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004924073737049784,
      "loss": 4.9969,
      "mean_token_accuracy": 0.22181894928216933,
      "num_tokens": 21904951.0,
      "step": 9550
    },
    {
      "entropy": 5.275024127960205,
      "epoch": 0.9178674351585014,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004923985401681528,
      "loss": 5.0291,
      "mean_token_accuracy": 0.22066223621368408,
      "num_tokens": 21917100.0,
      "step": 9555
    },
    {
      "entropy": 5.337822961807251,
      "epoch": 0.9183477425552353,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004923897015839788,
      "loss": 5.0538,
      "mean_token_accuracy": 0.21626515835523605,
      "num_tokens": 21927541.0,
      "step": 9560
    },
    {
      "entropy": 5.336556720733642,
      "epoch": 0.9188280499519692,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004923808579526613,
      "loss": 5.0086,
      "mean_token_accuracy": 0.21339127868413926,
      "num_tokens": 21938723.0,
      "step": 9565
    },
    {
      "entropy": 5.2254979610443115,
      "epoch": 0.9193083573487032,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004923720092744059,
      "loss": 4.9182,
      "mean_token_accuracy": 0.22439006417989732,
      "num_tokens": 21950424.0,
      "step": 9570
    },
    {
      "entropy": 5.319031810760498,
      "epoch": 0.9197886647454371,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004923631555494179,
      "loss": 5.1,
      "mean_token_accuracy": 0.21599253118038178,
      "num_tokens": 21961030.0,
      "step": 9575
    },
    {
      "entropy": 5.3590456485748295,
      "epoch": 0.920268972142171,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004923542967779028,
      "loss": 4.992,
      "mean_token_accuracy": 0.22134884148836137,
      "num_tokens": 21971625.0,
      "step": 9580
    },
    {
      "entropy": 5.331766605377197,
      "epoch": 0.920749279538905,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004923454329600664,
      "loss": 5.02,
      "mean_token_accuracy": 0.22271307110786437,
      "num_tokens": 21983733.0,
      "step": 9585
    },
    {
      "entropy": 5.21148910522461,
      "epoch": 0.9212295869356388,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004923365640961143,
      "loss": 4.959,
      "mean_token_accuracy": 0.22414645403623581,
      "num_tokens": 21995621.0,
      "step": 9590
    },
    {
      "entropy": 5.326543617248535,
      "epoch": 0.9217098943323727,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004923276901862526,
      "loss": 5.0457,
      "mean_token_accuracy": 0.21492437720298768,
      "num_tokens": 22007325.0,
      "step": 9595
    },
    {
      "entropy": 5.338121604919434,
      "epoch": 0.9221902017291066,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004923188112306874,
      "loss": 5.013,
      "mean_token_accuracy": 0.21119545251131058,
      "num_tokens": 22017733.0,
      "step": 9600
    },
    {
      "entropy": 5.311709928512573,
      "epoch": 0.9226705091258406,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004923099272296246,
      "loss": 5.0929,
      "mean_token_accuracy": 0.2158830925822258,
      "num_tokens": 22030451.0,
      "step": 9605
    },
    {
      "entropy": 5.302217769622803,
      "epoch": 0.9231508165225745,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004923010381832706,
      "loss": 5.0647,
      "mean_token_accuracy": 0.2070835277438164,
      "num_tokens": 22042626.0,
      "step": 9610
    },
    {
      "entropy": 5.296425485610962,
      "epoch": 0.9236311239193083,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004922921440918318,
      "loss": 5.0409,
      "mean_token_accuracy": 0.2172937884926796,
      "num_tokens": 22053314.0,
      "step": 9615
    },
    {
      "entropy": 5.410177755355835,
      "epoch": 0.9241114313160422,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004922832449555144,
      "loss": 5.134,
      "mean_token_accuracy": 0.2054891139268875,
      "num_tokens": 22064395.0,
      "step": 9620
    },
    {
      "entropy": 5.317427635192871,
      "epoch": 0.9245917387127762,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004922743407745255,
      "loss": 4.9991,
      "mean_token_accuracy": 0.21080311834812165,
      "num_tokens": 22075960.0,
      "step": 9625
    },
    {
      "entropy": 5.362533092498779,
      "epoch": 0.9250720461095101,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004922654315490714,
      "loss": 5.0941,
      "mean_token_accuracy": 0.2113686829805374,
      "num_tokens": 22086034.0,
      "step": 9630
    },
    {
      "entropy": 5.32300591468811,
      "epoch": 0.925552353506244,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004922565172793593,
      "loss": 5.1067,
      "mean_token_accuracy": 0.21166470497846604,
      "num_tokens": 22096184.0,
      "step": 9635
    },
    {
      "entropy": 5.318655776977539,
      "epoch": 0.9260326609029779,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004922475979655958,
      "loss": 5.062,
      "mean_token_accuracy": 0.21166544556617736,
      "num_tokens": 22108795.0,
      "step": 9640
    },
    {
      "entropy": 5.308929347991944,
      "epoch": 0.9265129682997119,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004922386736079883,
      "loss": 5.0316,
      "mean_token_accuracy": 0.21509423106908798,
      "num_tokens": 22119608.0,
      "step": 9645
    },
    {
      "entropy": 5.282072877883911,
      "epoch": 0.9269932756964457,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004922297442067438,
      "loss": 5.0068,
      "mean_token_accuracy": 0.22809043079614638,
      "num_tokens": 22131621.0,
      "step": 9650
    },
    {
      "entropy": 5.35063099861145,
      "epoch": 0.9274735830931796,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004922208097620697,
      "loss": 4.9682,
      "mean_token_accuracy": 0.21961411088705063,
      "num_tokens": 22142745.0,
      "step": 9655
    },
    {
      "entropy": 5.288157415390015,
      "epoch": 0.9279538904899135,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004922118702741735,
      "loss": 5.1706,
      "mean_token_accuracy": 0.21045637726783753,
      "num_tokens": 22155457.0,
      "step": 9660
    },
    {
      "entropy": 5.376977062225341,
      "epoch": 0.9284341978866475,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004922029257432625,
      "loss": 5.039,
      "mean_token_accuracy": 0.226725772023201,
      "num_tokens": 22165955.0,
      "step": 9665
    },
    {
      "entropy": 5.312184190750122,
      "epoch": 0.9289145052833814,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004921939761695446,
      "loss": 4.9942,
      "mean_token_accuracy": 0.2166220709681511,
      "num_tokens": 22178142.0,
      "step": 9670
    },
    {
      "entropy": 5.200202417373657,
      "epoch": 0.9293948126801153,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004921850215532275,
      "loss": 4.9521,
      "mean_token_accuracy": 0.22166887521743775,
      "num_tokens": 22190315.0,
      "step": 9675
    },
    {
      "entropy": 5.2593049049377445,
      "epoch": 0.9298751200768491,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004921760618945192,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21394742131233216,
      "num_tokens": 22201785.0,
      "step": 9680
    },
    {
      "entropy": 5.352148580551147,
      "epoch": 0.9303554274735831,
      "grad_norm": 1.0,
      "learning_rate": 0.0004921670971936276,
      "loss": 4.9955,
      "mean_token_accuracy": 0.21628571152687073,
      "num_tokens": 22212471.0,
      "step": 9685
    },
    {
      "entropy": 5.338489437103272,
      "epoch": 0.930835734870317,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004921581274507607,
      "loss": 5.0658,
      "mean_token_accuracy": 0.21684152036905288,
      "num_tokens": 22223188.0,
      "step": 9690
    },
    {
      "entropy": 5.256138706207276,
      "epoch": 0.9313160422670509,
      "grad_norm": 1.015625,
      "learning_rate": 0.000492149152666127,
      "loss": 5.0533,
      "mean_token_accuracy": 0.21721217930316924,
      "num_tokens": 22233978.0,
      "step": 9695
    },
    {
      "entropy": 5.289171981811523,
      "epoch": 0.9317963496637848,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004921401728399348,
      "loss": 5.021,
      "mean_token_accuracy": 0.22602349668741226,
      "num_tokens": 22244713.0,
      "step": 9700
    },
    {
      "entropy": 5.324395084381104,
      "epoch": 0.9322766570605188,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004921311879723926,
      "loss": 5.0602,
      "mean_token_accuracy": 0.21518180519342422,
      "num_tokens": 22256192.0,
      "step": 9705
    },
    {
      "entropy": 5.344833564758301,
      "epoch": 0.9327569644572526,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004921221980637088,
      "loss": 5.032,
      "mean_token_accuracy": 0.21172490566968918,
      "num_tokens": 22268294.0,
      "step": 9710
    },
    {
      "entropy": 5.329265499114991,
      "epoch": 0.9332372718539865,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004921132031140925,
      "loss": 5.0707,
      "mean_token_accuracy": 0.2170101523399353,
      "num_tokens": 22278952.0,
      "step": 9715
    },
    {
      "entropy": 5.308583831787109,
      "epoch": 0.9337175792507204,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004921042031237521,
      "loss": 5.014,
      "mean_token_accuracy": 0.21303013563156128,
      "num_tokens": 22291057.0,
      "step": 9720
    },
    {
      "entropy": 5.380600261688232,
      "epoch": 0.9341978866474544,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004920951980928969,
      "loss": 5.1008,
      "mean_token_accuracy": 0.2134473741054535,
      "num_tokens": 22302479.0,
      "step": 9725
    },
    {
      "entropy": 5.393728160858155,
      "epoch": 0.9346781940441883,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004920861880217359,
      "loss": 5.1576,
      "mean_token_accuracy": 0.2076900690793991,
      "num_tokens": 22315116.0,
      "step": 9730
    },
    {
      "entropy": 5.346706485748291,
      "epoch": 0.9351585014409222,
      "grad_norm": 0.875,
      "learning_rate": 0.0004920771729104781,
      "loss": 5.0816,
      "mean_token_accuracy": 0.21419650614261626,
      "num_tokens": 22327548.0,
      "step": 9735
    },
    {
      "entropy": 5.321560192108154,
      "epoch": 0.9356388088376562,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004920681527593329,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21998959481716157,
      "num_tokens": 22339154.0,
      "step": 9740
    },
    {
      "entropy": 5.282303524017334,
      "epoch": 0.93611911623439,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004920591275685098,
      "loss": 5.0665,
      "mean_token_accuracy": 0.21839868724346162,
      "num_tokens": 22350781.0,
      "step": 9745
    },
    {
      "entropy": 5.356006240844726,
      "epoch": 0.9365994236311239,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004920500973382184,
      "loss": 5.0666,
      "mean_token_accuracy": 0.21905728429555893,
      "num_tokens": 22361990.0,
      "step": 9750
    },
    {
      "entropy": 5.301377773284912,
      "epoch": 0.9370797310278578,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004920410620686682,
      "loss": 4.9457,
      "mean_token_accuracy": 0.22767526805400848,
      "num_tokens": 22372973.0,
      "step": 9755
    },
    {
      "entropy": 5.227727270126342,
      "epoch": 0.9375600384245918,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004920320217600689,
      "loss": 4.9602,
      "mean_token_accuracy": 0.22442955523729324,
      "num_tokens": 22384369.0,
      "step": 9760
    },
    {
      "entropy": 5.319287061691284,
      "epoch": 0.9380403458213257,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004920229764126306,
      "loss": 5.0662,
      "mean_token_accuracy": 0.22147373408079146,
      "num_tokens": 22395792.0,
      "step": 9765
    },
    {
      "entropy": 5.307302093505859,
      "epoch": 0.9385206532180596,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004920139260265632,
      "loss": 5.0485,
      "mean_token_accuracy": 0.21197786629199983,
      "num_tokens": 22408182.0,
      "step": 9770
    },
    {
      "entropy": 5.41457166671753,
      "epoch": 0.9390009606147934,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004920048706020769,
      "loss": 5.1802,
      "mean_token_accuracy": 0.2068299189209938,
      "num_tokens": 22419774.0,
      "step": 9775
    },
    {
      "entropy": 5.232258939743042,
      "epoch": 0.9394812680115274,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004919958101393817,
      "loss": 4.8999,
      "mean_token_accuracy": 0.2266707479953766,
      "num_tokens": 22430210.0,
      "step": 9780
    },
    {
      "entropy": 5.311747121810913,
      "epoch": 0.9399615754082613,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004919867446386883,
      "loss": 5.0498,
      "mean_token_accuracy": 0.217346715927124,
      "num_tokens": 22442444.0,
      "step": 9785
    },
    {
      "entropy": 5.247062921524048,
      "epoch": 0.9404418828049952,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000491977674100207,
      "loss": 4.9936,
      "mean_token_accuracy": 0.22415463924407958,
      "num_tokens": 22455521.0,
      "step": 9790
    },
    {
      "entropy": 5.370890474319458,
      "epoch": 0.9409221902017291,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004919685985241483,
      "loss": 5.0677,
      "mean_token_accuracy": 0.21972337514162063,
      "num_tokens": 22466997.0,
      "step": 9795
    },
    {
      "entropy": 5.3071897506713865,
      "epoch": 0.9414024975984631,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000491959517910723,
      "loss": 4.9819,
      "mean_token_accuracy": 0.2227747544646263,
      "num_tokens": 22477851.0,
      "step": 9800
    },
    {
      "entropy": 5.238428831100464,
      "epoch": 0.9418828049951969,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004919504322601421,
      "loss": 5.0921,
      "mean_token_accuracy": 0.21362376660108567,
      "num_tokens": 22489319.0,
      "step": 9805
    },
    {
      "entropy": 5.2027193069458,
      "epoch": 0.9423631123919308,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004919413415726162,
      "loss": 4.9467,
      "mean_token_accuracy": 0.22427940219640732,
      "num_tokens": 22500847.0,
      "step": 9810
    },
    {
      "entropy": 5.342273426055908,
      "epoch": 0.9428434197886647,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004919322458483566,
      "loss": 5.0326,
      "mean_token_accuracy": 0.21770778596401213,
      "num_tokens": 22512719.0,
      "step": 9815
    },
    {
      "entropy": 5.254771661758423,
      "epoch": 0.9433237271853987,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004919231450875745,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22413523644208908,
      "num_tokens": 22522984.0,
      "step": 9820
    },
    {
      "entropy": 5.284532117843628,
      "epoch": 0.9438040345821326,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004919140392904809,
      "loss": 5.0201,
      "mean_token_accuracy": 0.2202260047197342,
      "num_tokens": 22534816.0,
      "step": 9825
    },
    {
      "entropy": 5.298043966293335,
      "epoch": 0.9442843419788665,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004919049284572875,
      "loss": 5.0092,
      "mean_token_accuracy": 0.21470444798469543,
      "num_tokens": 22545753.0,
      "step": 9830
    },
    {
      "entropy": 5.342241621017456,
      "epoch": 0.9447646493756003,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004918958125882058,
      "loss": 5.0906,
      "mean_token_accuracy": 0.21063587814569473,
      "num_tokens": 22557237.0,
      "step": 9835
    },
    {
      "entropy": 5.33295259475708,
      "epoch": 0.9452449567723343,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004918866916834474,
      "loss": 5.0021,
      "mean_token_accuracy": 0.22219055742025376,
      "num_tokens": 22568909.0,
      "step": 9840
    },
    {
      "entropy": 5.391683340072632,
      "epoch": 0.9457252641690682,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004918775657432239,
      "loss": 5.1479,
      "mean_token_accuracy": 0.2095224678516388,
      "num_tokens": 22582162.0,
      "step": 9845
    },
    {
      "entropy": 5.2753864288330075,
      "epoch": 0.9462055715658021,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004918684347677474,
      "loss": 5.0134,
      "mean_token_accuracy": 0.21508377343416213,
      "num_tokens": 22592405.0,
      "step": 9850
    },
    {
      "entropy": 5.312293910980225,
      "epoch": 0.946685878962536,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004918592987572298,
      "loss": 5.0334,
      "mean_token_accuracy": 0.21834415048360825,
      "num_tokens": 22603588.0,
      "step": 9855
    },
    {
      "entropy": 5.402937126159668,
      "epoch": 0.94716618635927,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004918501577118832,
      "loss": 5.1031,
      "mean_token_accuracy": 0.21975343078374862,
      "num_tokens": 22614995.0,
      "step": 9860
    },
    {
      "entropy": 5.321914148330689,
      "epoch": 0.9476464937560038,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00049184101163192,
      "loss": 5.0575,
      "mean_token_accuracy": 0.21013206094503403,
      "num_tokens": 22627556.0,
      "step": 9865
    },
    {
      "entropy": 5.3679169654846195,
      "epoch": 0.9481268011527377,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004918318605175522,
      "loss": 5.0439,
      "mean_token_accuracy": 0.21720885038375853,
      "num_tokens": 22638339.0,
      "step": 9870
    },
    {
      "entropy": 5.350629711151123,
      "epoch": 0.9486071085494716,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004918227043689924,
      "loss": 4.9991,
      "mean_token_accuracy": 0.2213841661810875,
      "num_tokens": 22648922.0,
      "step": 9875
    },
    {
      "entropy": 5.2442409038543705,
      "epoch": 0.9490874159462056,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004918135431864534,
      "loss": 5.032,
      "mean_token_accuracy": 0.22041068822145463,
      "num_tokens": 22662006.0,
      "step": 9880
    },
    {
      "entropy": 5.247269010543823,
      "epoch": 0.9495677233429395,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004918043769701478,
      "loss": 4.9911,
      "mean_token_accuracy": 0.22019424736499787,
      "num_tokens": 22674649.0,
      "step": 9885
    },
    {
      "entropy": 5.4095478534698485,
      "epoch": 0.9500480307396734,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004917952057202882,
      "loss": 5.0953,
      "mean_token_accuracy": 0.20772865414619446,
      "num_tokens": 22685971.0,
      "step": 9890
    },
    {
      "entropy": 5.276963806152343,
      "epoch": 0.9505283381364072,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004917860294370877,
      "loss": 4.9559,
      "mean_token_accuracy": 0.22325167655944825,
      "num_tokens": 22696174.0,
      "step": 9895
    },
    {
      "entropy": 5.352025890350342,
      "epoch": 0.9510086455331412,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004917768481207593,
      "loss": 5.0615,
      "mean_token_accuracy": 0.2159486547112465,
      "num_tokens": 22706983.0,
      "step": 9900
    },
    {
      "entropy": 5.306163692474366,
      "epoch": 0.9514889529298751,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004917676617715162,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21577799916267396,
      "num_tokens": 22718251.0,
      "step": 9905
    },
    {
      "entropy": 5.232373142242432,
      "epoch": 0.951969260326609,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004917584703895717,
      "loss": 4.9753,
      "mean_token_accuracy": 0.2236893966794014,
      "num_tokens": 22730071.0,
      "step": 9910
    },
    {
      "entropy": 5.301542043685913,
      "epoch": 0.952449567723343,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004917492739751391,
      "loss": 5.0293,
      "mean_token_accuracy": 0.2162534102797508,
      "num_tokens": 22742027.0,
      "step": 9915
    },
    {
      "entropy": 5.254919338226318,
      "epoch": 0.9529298751200769,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000491740072528432,
      "loss": 4.921,
      "mean_token_accuracy": 0.23174448013305665,
      "num_tokens": 22752946.0,
      "step": 9920
    },
    {
      "entropy": 5.308495092391968,
      "epoch": 0.9534101825168108,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000491730866049664,
      "loss": 5.0065,
      "mean_token_accuracy": 0.21784865111112595,
      "num_tokens": 22763944.0,
      "step": 9925
    },
    {
      "entropy": 5.2533105373382565,
      "epoch": 0.9538904899135446,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004917216545390489,
      "loss": 4.9875,
      "mean_token_accuracy": 0.2221623733639717,
      "num_tokens": 22774414.0,
      "step": 9930
    },
    {
      "entropy": 5.23083815574646,
      "epoch": 0.9543707973102786,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004917124379968004,
      "loss": 4.9224,
      "mean_token_accuracy": 0.229240345954895,
      "num_tokens": 22785533.0,
      "step": 9935
    },
    {
      "entropy": 5.204967308044433,
      "epoch": 0.9548511047070125,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004917032164231327,
      "loss": 4.8796,
      "mean_token_accuracy": 0.22522834241390227,
      "num_tokens": 22795809.0,
      "step": 9940
    },
    {
      "entropy": 5.280347633361816,
      "epoch": 0.9553314121037464,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004916939898182598,
      "loss": 5.1031,
      "mean_token_accuracy": 0.212029105424881,
      "num_tokens": 22807705.0,
      "step": 9945
    },
    {
      "entropy": 5.3249047756195065,
      "epoch": 0.9558117195004803,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004916847581823958,
      "loss": 5.0117,
      "mean_token_accuracy": 0.21814230233430862,
      "num_tokens": 22818852.0,
      "step": 9950
    },
    {
      "entropy": 5.259893798828125,
      "epoch": 0.9562920268972143,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004916755215157552,
      "loss": 4.9319,
      "mean_token_accuracy": 0.22051620334386826,
      "num_tokens": 22829146.0,
      "step": 9955
    },
    {
      "entropy": 5.267180728912353,
      "epoch": 0.9567723342939481,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004916662798185524,
      "loss": 5.0069,
      "mean_token_accuracy": 0.22495510578155517,
      "num_tokens": 22840088.0,
      "step": 9960
    },
    {
      "entropy": 5.27195463180542,
      "epoch": 0.957252641690682,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004916570330910019,
      "loss": 5.035,
      "mean_token_accuracy": 0.21801164746284485,
      "num_tokens": 22852470.0,
      "step": 9965
    },
    {
      "entropy": 5.344923496246338,
      "epoch": 0.9577329490874159,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004916477813333185,
      "loss": 5.0607,
      "mean_token_accuracy": 0.2133091852068901,
      "num_tokens": 22863673.0,
      "step": 9970
    },
    {
      "entropy": 5.307507467269898,
      "epoch": 0.9582132564841499,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004916385245457168,
      "loss": 5.0453,
      "mean_token_accuracy": 0.2134061262011528,
      "num_tokens": 22874888.0,
      "step": 9975
    },
    {
      "entropy": 5.28753137588501,
      "epoch": 0.9586935638808838,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000491629262728412,
      "loss": 5.0609,
      "mean_token_accuracy": 0.2115960195660591,
      "num_tokens": 22886811.0,
      "step": 9980
    },
    {
      "entropy": 5.3310610294342045,
      "epoch": 0.9591738712776177,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004916199958816188,
      "loss": 5.016,
      "mean_token_accuracy": 0.21616479456424714,
      "num_tokens": 22898777.0,
      "step": 9985
    },
    {
      "entropy": 5.311865901947021,
      "epoch": 0.9596541786743515,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004916107240055527,
      "loss": 4.9896,
      "mean_token_accuracy": 0.22230456322431563,
      "num_tokens": 22910804.0,
      "step": 9990
    },
    {
      "entropy": 5.363910245895386,
      "epoch": 0.9601344860710855,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004916014471004287,
      "loss": 5.1255,
      "mean_token_accuracy": 0.22072769999504088,
      "num_tokens": 22922002.0,
      "step": 9995
    },
    {
      "entropy": 5.326347064971924,
      "epoch": 0.9606147934678194,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915921651664622,
      "loss": 5.0207,
      "mean_token_accuracy": 0.21681399047374725,
      "num_tokens": 22933471.0,
      "step": 10000
    },
    {
      "entropy": 5.217788887023926,
      "epoch": 0.9610951008645533,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000491582878203869,
      "loss": 4.9353,
      "mean_token_accuracy": 0.22138506919145584,
      "num_tokens": 22945303.0,
      "step": 10005
    },
    {
      "entropy": 5.301219272613525,
      "epoch": 0.9615754082612872,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004915735862128643,
      "loss": 5.0797,
      "mean_token_accuracy": 0.206269071996212,
      "num_tokens": 22956620.0,
      "step": 10010
    },
    {
      "entropy": 5.353693246841431,
      "epoch": 0.9620557156580212,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004915642891936641,
      "loss": 5.057,
      "mean_token_accuracy": 0.21933349668979646,
      "num_tokens": 22968941.0,
      "step": 10015
    },
    {
      "entropy": 5.334658861160278,
      "epoch": 0.962536023054755,
      "grad_norm": 1.0,
      "learning_rate": 0.0004915549871464841,
      "loss": 5.031,
      "mean_token_accuracy": 0.21822743117809296,
      "num_tokens": 22980222.0,
      "step": 10020
    },
    {
      "entropy": 5.399587202072143,
      "epoch": 0.9630163304514889,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004915456800715403,
      "loss": 5.0572,
      "mean_token_accuracy": 0.21029905676841737,
      "num_tokens": 22991156.0,
      "step": 10025
    },
    {
      "entropy": 5.378714990615845,
      "epoch": 0.9634966378482228,
      "grad_norm": 0.984375,
      "learning_rate": 0.000491536367969049,
      "loss": 5.1943,
      "mean_token_accuracy": 0.20684890300035477,
      "num_tokens": 23002939.0,
      "step": 10030
    },
    {
      "entropy": 5.394938707351685,
      "epoch": 0.9639769452449568,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004915270508392261,
      "loss": 5.0409,
      "mean_token_accuracy": 0.2146536812186241,
      "num_tokens": 23015590.0,
      "step": 10035
    },
    {
      "entropy": 5.225053167343139,
      "epoch": 0.9644572526416907,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000491517728682288,
      "loss": 4.9237,
      "mean_token_accuracy": 0.2257880300283432,
      "num_tokens": 23026387.0,
      "step": 10040
    },
    {
      "entropy": 5.200448131561279,
      "epoch": 0.9649375600384246,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004915084014984512,
      "loss": 4.9942,
      "mean_token_accuracy": 0.22349174916744233,
      "num_tokens": 23037475.0,
      "step": 10045
    },
    {
      "entropy": 5.275051689147949,
      "epoch": 0.9654178674351584,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004914990692879322,
      "loss": 4.9703,
      "mean_token_accuracy": 0.22475121766328812,
      "num_tokens": 23049305.0,
      "step": 10050
    },
    {
      "entropy": 5.295267486572266,
      "epoch": 0.9658981748318924,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004914897320509478,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21849175542593002,
      "num_tokens": 23061765.0,
      "step": 10055
    },
    {
      "entropy": 5.263977956771851,
      "epoch": 0.9663784822286263,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004914803897877146,
      "loss": 4.9895,
      "mean_token_accuracy": 0.21232639402151107,
      "num_tokens": 23072355.0,
      "step": 10060
    },
    {
      "entropy": 5.322015285491943,
      "epoch": 0.9668587896253602,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004914710424984495,
      "loss": 5.0159,
      "mean_token_accuracy": 0.21301655173301698,
      "num_tokens": 23085583.0,
      "step": 10065
    },
    {
      "entropy": 5.359940147399902,
      "epoch": 0.9673390970220941,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004914616901833696,
      "loss": 5.0569,
      "mean_token_accuracy": 0.21371917724609374,
      "num_tokens": 23095942.0,
      "step": 10070
    },
    {
      "entropy": 5.290450143814087,
      "epoch": 0.9678194044188281,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491452332842692,
      "loss": 4.9781,
      "mean_token_accuracy": 0.22154778093099595,
      "num_tokens": 23106540.0,
      "step": 10075
    },
    {
      "entropy": 5.289186191558838,
      "epoch": 0.968299711815562,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000491442970476634,
      "loss": 5.0465,
      "mean_token_accuracy": 0.2206612914800644,
      "num_tokens": 23118006.0,
      "step": 10080
    },
    {
      "entropy": 5.276647758483887,
      "epoch": 0.9687800192122958,
      "grad_norm": 1.0,
      "learning_rate": 0.0004914336030854129,
      "loss": 4.9358,
      "mean_token_accuracy": 0.23133217096328734,
      "num_tokens": 23129103.0,
      "step": 10085
    },
    {
      "entropy": 5.241673278808594,
      "epoch": 0.9692603266090298,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004914242306692461,
      "loss": 4.9537,
      "mean_token_accuracy": 0.22525297850370407,
      "num_tokens": 23140009.0,
      "step": 10090
    },
    {
      "entropy": 5.271966123580933,
      "epoch": 0.9697406340057637,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004914148532283516,
      "loss": 5.0058,
      "mean_token_accuracy": 0.22268082499504088,
      "num_tokens": 23150982.0,
      "step": 10095
    },
    {
      "entropy": 5.319319438934326,
      "epoch": 0.9702209414024976,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004914054707629466,
      "loss": 5.0186,
      "mean_token_accuracy": 0.21841463893651963,
      "num_tokens": 23161834.0,
      "step": 10100
    },
    {
      "entropy": 5.315860605239868,
      "epoch": 0.9707012487992315,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004913960832732493,
      "loss": 5.1551,
      "mean_token_accuracy": 0.2086580440402031,
      "num_tokens": 23173355.0,
      "step": 10105
    },
    {
      "entropy": 5.340712308883667,
      "epoch": 0.9711815561959655,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004913866907594774,
      "loss": 5.077,
      "mean_token_accuracy": 0.2130504235625267,
      "num_tokens": 23185075.0,
      "step": 10110
    },
    {
      "entropy": 5.310555458068848,
      "epoch": 0.9716618635926993,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004913772932218491,
      "loss": 5.0368,
      "mean_token_accuracy": 0.22630117386579512,
      "num_tokens": 23195590.0,
      "step": 10115
    },
    {
      "entropy": 5.365739870071411,
      "epoch": 0.9721421709894332,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004913678906605825,
      "loss": 5.0075,
      "mean_token_accuracy": 0.22077048271894456,
      "num_tokens": 23207668.0,
      "step": 10120
    },
    {
      "entropy": 5.23333625793457,
      "epoch": 0.9726224783861671,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004913584830758961,
      "loss": 5.0027,
      "mean_token_accuracy": 0.2244031935930252,
      "num_tokens": 23218497.0,
      "step": 10125
    },
    {
      "entropy": 5.389020872116089,
      "epoch": 0.9731027857829011,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004913490704680081,
      "loss": 5.0666,
      "mean_token_accuracy": 0.2115566238760948,
      "num_tokens": 23230575.0,
      "step": 10130
    },
    {
      "entropy": 5.346340274810791,
      "epoch": 0.973583093179635,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004913396528371371,
      "loss": 5.0684,
      "mean_token_accuracy": 0.21991017907857896,
      "num_tokens": 23242348.0,
      "step": 10135
    },
    {
      "entropy": 5.30429105758667,
      "epoch": 0.9740634005763689,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004913302301835018,
      "loss": 5.0397,
      "mean_token_accuracy": 0.2227107137441635,
      "num_tokens": 23253297.0,
      "step": 10140
    },
    {
      "entropy": 5.235578298568726,
      "epoch": 0.9745437079731027,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000491320802507321,
      "loss": 4.9828,
      "mean_token_accuracy": 0.22600618600845337,
      "num_tokens": 23265830.0,
      "step": 10145
    },
    {
      "entropy": 5.336393928527832,
      "epoch": 0.9750240153698367,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004913113698088133,
      "loss": 5.0716,
      "mean_token_accuracy": 0.21149585098028184,
      "num_tokens": 23275591.0,
      "step": 10150
    },
    {
      "entropy": 5.292669630050659,
      "epoch": 0.9755043227665706,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000491301932088198,
      "loss": 4.9589,
      "mean_token_accuracy": 0.22035377323627472,
      "num_tokens": 23286685.0,
      "step": 10155
    },
    {
      "entropy": 5.245239114761352,
      "epoch": 0.9759846301633045,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004912924893456942,
      "loss": 4.9662,
      "mean_token_accuracy": 0.22627130299806594,
      "num_tokens": 23298776.0,
      "step": 10160
    },
    {
      "entropy": 5.2822143077850345,
      "epoch": 0.9764649375600384,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000491283041581521,
      "loss": 4.9553,
      "mean_token_accuracy": 0.22000742107629775,
      "num_tokens": 23310498.0,
      "step": 10165
    },
    {
      "entropy": 5.3116693019866945,
      "epoch": 0.9769452449567724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004912735887958978,
      "loss": 5.0149,
      "mean_token_accuracy": 0.21783301830291749,
      "num_tokens": 23321089.0,
      "step": 10170
    },
    {
      "entropy": 5.30176157951355,
      "epoch": 0.9774255523535063,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004912641309890441,
      "loss": 4.9953,
      "mean_token_accuracy": 0.21955489069223405,
      "num_tokens": 23332142.0,
      "step": 10175
    },
    {
      "entropy": 5.249406337738037,
      "epoch": 0.9779058597502401,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004912546681611794,
      "loss": 4.9659,
      "mean_token_accuracy": 0.22600394636392593,
      "num_tokens": 23343014.0,
      "step": 10180
    },
    {
      "entropy": 5.280077028274536,
      "epoch": 0.978386167146974,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004912452003125234,
      "loss": 4.9434,
      "mean_token_accuracy": 0.22357450872659684,
      "num_tokens": 23354611.0,
      "step": 10185
    },
    {
      "entropy": 5.28184928894043,
      "epoch": 0.978866474543708,
      "grad_norm": 0.9375,
      "learning_rate": 0.000491235727443296,
      "loss": 5.0028,
      "mean_token_accuracy": 0.2151727020740509,
      "num_tokens": 23365608.0,
      "step": 10190
    },
    {
      "entropy": 5.333352279663086,
      "epoch": 0.9793467819404419,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004912262495537171,
      "loss": 5.0342,
      "mean_token_accuracy": 0.22038183957338334,
      "num_tokens": 23377884.0,
      "step": 10195
    },
    {
      "entropy": 5.221385526657104,
      "epoch": 0.9798270893371758,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004912167666440068,
      "loss": 4.9425,
      "mean_token_accuracy": 0.22146084159612656,
      "num_tokens": 23389553.0,
      "step": 10200
    },
    {
      "entropy": 5.25901231765747,
      "epoch": 0.9803073967339097,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004912072787143852,
      "loss": 4.9284,
      "mean_token_accuracy": 0.22336874306201934,
      "num_tokens": 23401079.0,
      "step": 10205
    },
    {
      "entropy": 5.230851697921753,
      "epoch": 0.9807877041306436,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004911977857650725,
      "loss": 5.001,
      "mean_token_accuracy": 0.21796642541885375,
      "num_tokens": 23412886.0,
      "step": 10210
    },
    {
      "entropy": 5.313813924789429,
      "epoch": 0.9812680115273775,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004911882877962893,
      "loss": 5.05,
      "mean_token_accuracy": 0.21360267251729964,
      "num_tokens": 23424758.0,
      "step": 10215
    },
    {
      "entropy": 5.253488969802857,
      "epoch": 0.9817483189241114,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004911787848082559,
      "loss": 4.9236,
      "mean_token_accuracy": 0.23152619004249572,
      "num_tokens": 23435552.0,
      "step": 10220
    },
    {
      "entropy": 5.21381630897522,
      "epoch": 0.9822286263208453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911692768011931,
      "loss": 4.9279,
      "mean_token_accuracy": 0.22218966335058213,
      "num_tokens": 23446584.0,
      "step": 10225
    },
    {
      "entropy": 5.370476961135864,
      "epoch": 0.9827089337175793,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004911597637753217,
      "loss": 5.1431,
      "mean_token_accuracy": 0.20453443825244905,
      "num_tokens": 23458452.0,
      "step": 10230
    },
    {
      "entropy": 5.317557525634766,
      "epoch": 0.9831892411143132,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004911502457308623,
      "loss": 5.0181,
      "mean_token_accuracy": 0.21623304039239882,
      "num_tokens": 23470310.0,
      "step": 10235
    },
    {
      "entropy": 5.359401750564575,
      "epoch": 0.983669548511047,
      "grad_norm": 1.0,
      "learning_rate": 0.000491140722668036,
      "loss": 5.0556,
      "mean_token_accuracy": 0.2201622560620308,
      "num_tokens": 23481166.0,
      "step": 10240
    },
    {
      "entropy": 5.263864898681641,
      "epoch": 0.984149855907781,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000491131194587064,
      "loss": 5.05,
      "mean_token_accuracy": 0.22096795290708543,
      "num_tokens": 23493134.0,
      "step": 10245
    },
    {
      "entropy": 5.236112356185913,
      "epoch": 0.9846301633045149,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004911216614881675,
      "loss": 4.9234,
      "mean_token_accuracy": 0.22464846521615983,
      "num_tokens": 23504983.0,
      "step": 10250
    },
    {
      "entropy": 5.36480712890625,
      "epoch": 0.9851104707012488,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004911121233715677,
      "loss": 5.0134,
      "mean_token_accuracy": 0.21945274621248245,
      "num_tokens": 23516119.0,
      "step": 10255
    },
    {
      "entropy": 5.3713061809539795,
      "epoch": 0.9855907780979827,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004911025802374861,
      "loss": 5.0491,
      "mean_token_accuracy": 0.21678780764341354,
      "num_tokens": 23528242.0,
      "step": 10260
    },
    {
      "entropy": 5.261427736282348,
      "epoch": 0.9860710854947167,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004910930320861442,
      "loss": 4.9434,
      "mean_token_accuracy": 0.22971244156360626,
      "num_tokens": 23539738.0,
      "step": 10265
    },
    {
      "entropy": 5.139218854904175,
      "epoch": 0.9865513928914506,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004910834789177639,
      "loss": 4.9503,
      "mean_token_accuracy": 0.22372818887233734,
      "num_tokens": 23551228.0,
      "step": 10270
    },
    {
      "entropy": 5.296945858001709,
      "epoch": 0.9870317002881844,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004910739207325668,
      "loss": 5.0086,
      "mean_token_accuracy": 0.22216663062572478,
      "num_tokens": 23563084.0,
      "step": 10275
    },
    {
      "entropy": 5.269334220886231,
      "epoch": 0.9875120076849183,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004910643575307749,
      "loss": 4.9696,
      "mean_token_accuracy": 0.22117464244365692,
      "num_tokens": 23574328.0,
      "step": 10280
    },
    {
      "entropy": 5.239086723327636,
      "epoch": 0.9879923150816523,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004910547893126102,
      "loss": 4.9634,
      "mean_token_accuracy": 0.22538317143917083,
      "num_tokens": 23585230.0,
      "step": 10285
    },
    {
      "entropy": 5.317661619186401,
      "epoch": 0.9884726224783862,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004910452160782948,
      "loss": 4.9957,
      "mean_token_accuracy": 0.21143541485071182,
      "num_tokens": 23596951.0,
      "step": 10290
    },
    {
      "entropy": 5.322635316848755,
      "epoch": 0.9889529298751201,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000491035637828051,
      "loss": 5.0239,
      "mean_token_accuracy": 0.2191610798239708,
      "num_tokens": 23607759.0,
      "step": 10295
    },
    {
      "entropy": 5.2963786125183105,
      "epoch": 0.989433237271854,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004910260545621012,
      "loss": 5.038,
      "mean_token_accuracy": 0.2118114709854126,
      "num_tokens": 23619631.0,
      "step": 10300
    },
    {
      "entropy": 5.283452558517456,
      "epoch": 0.9899135446685879,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004910164662806679,
      "loss": 5.0385,
      "mean_token_accuracy": 0.2200574353337288,
      "num_tokens": 23630601.0,
      "step": 10305
    },
    {
      "entropy": 5.316462993621826,
      "epoch": 0.9903938520653218,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004910068729839736,
      "loss": 4.9624,
      "mean_token_accuracy": 0.22418494075536727,
      "num_tokens": 23641330.0,
      "step": 10310
    },
    {
      "entropy": 5.339647197723389,
      "epoch": 0.9908741594620557,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004909972746722413,
      "loss": 5.0432,
      "mean_token_accuracy": 0.21308202892541886,
      "num_tokens": 23651492.0,
      "step": 10315
    },
    {
      "entropy": 5.250365829467773,
      "epoch": 0.9913544668587896,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004909876713456935,
      "loss": 4.9377,
      "mean_token_accuracy": 0.22199930399656295,
      "num_tokens": 23661773.0,
      "step": 10320
    },
    {
      "entropy": 5.263142871856689,
      "epoch": 0.9918347742555236,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004909780630045534,
      "loss": 4.986,
      "mean_token_accuracy": 0.22256284654140474,
      "num_tokens": 23673534.0,
      "step": 10325
    },
    {
      "entropy": 5.32167158126831,
      "epoch": 0.9923150816522575,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000490968449649044,
      "loss": 5.0899,
      "mean_token_accuracy": 0.2141452521085739,
      "num_tokens": 23684892.0,
      "step": 10330
    },
    {
      "entropy": 5.3284827709198,
      "epoch": 0.9927953890489913,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004909588312793884,
      "loss": 5.0529,
      "mean_token_accuracy": 0.215768863260746,
      "num_tokens": 23696076.0,
      "step": 10335
    },
    {
      "entropy": 5.341275835037232,
      "epoch": 0.9932756964457252,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004909492078958101,
      "loss": 5.0501,
      "mean_token_accuracy": 0.21610995829105378,
      "num_tokens": 23707795.0,
      "step": 10340
    },
    {
      "entropy": 5.270211553573608,
      "epoch": 0.9937560038424592,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004909395794985324,
      "loss": 4.9528,
      "mean_token_accuracy": 0.2214967966079712,
      "num_tokens": 23720802.0,
      "step": 10345
    },
    {
      "entropy": 5.281257200241089,
      "epoch": 0.9942363112391931,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004909299460877788,
      "loss": 4.9859,
      "mean_token_accuracy": 0.21097991466522217,
      "num_tokens": 23732854.0,
      "step": 10350
    },
    {
      "entropy": 5.340284395217895,
      "epoch": 0.994716618635927,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004909203076637732,
      "loss": 5.0719,
      "mean_token_accuracy": 0.2112707808613777,
      "num_tokens": 23743593.0,
      "step": 10355
    },
    {
      "entropy": 5.335320615768433,
      "epoch": 0.9951969260326609,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004909106642267392,
      "loss": 5.0496,
      "mean_token_accuracy": 0.210444900393486,
      "num_tokens": 23755447.0,
      "step": 10360
    },
    {
      "entropy": 5.292027711868286,
      "epoch": 0.9956772334293948,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004909010157769006,
      "loss": 5.0494,
      "mean_token_accuracy": 0.2152780368924141,
      "num_tokens": 23767181.0,
      "step": 10365
    },
    {
      "entropy": 5.410845375061035,
      "epoch": 0.9961575408261287,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004908913623144814,
      "loss": 5.1146,
      "mean_token_accuracy": 0.207489912211895,
      "num_tokens": 23776356.0,
      "step": 10370
    },
    {
      "entropy": 5.2951267719268795,
      "epoch": 0.9966378482228626,
      "grad_norm": 1.0,
      "learning_rate": 0.000490881703839706,
      "loss": 5.071,
      "mean_token_accuracy": 0.21977514177560806,
      "num_tokens": 23787965.0,
      "step": 10375
    },
    {
      "entropy": 5.233155059814453,
      "epoch": 0.9971181556195965,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004908720403527984,
      "loss": 4.9308,
      "mean_token_accuracy": 0.2260810375213623,
      "num_tokens": 23800327.0,
      "step": 10380
    },
    {
      "entropy": 5.29593563079834,
      "epoch": 0.9975984630163305,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000490862371853983,
      "loss": 4.9414,
      "mean_token_accuracy": 0.21992311924695968,
      "num_tokens": 23812845.0,
      "step": 10385
    },
    {
      "entropy": 5.3876556873321535,
      "epoch": 0.9980787704130644,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004908526983434844,
      "loss": 5.1039,
      "mean_token_accuracy": 0.2107962265610695,
      "num_tokens": 23824831.0,
      "step": 10390
    },
    {
      "entropy": 5.280596160888672,
      "epoch": 0.9985590778097982,
      "grad_norm": 0.875,
      "learning_rate": 0.000490843019821527,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22450566440820693,
      "num_tokens": 23836697.0,
      "step": 10395
    },
    {
      "entropy": 5.263504076004028,
      "epoch": 0.9990393852065321,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004908333362883358,
      "loss": 4.9818,
      "mean_token_accuracy": 0.2218126595020294,
      "num_tokens": 23847112.0,
      "step": 10400
    },
    {
      "entropy": 5.309028577804566,
      "epoch": 0.9995196926032661,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004908236477441353,
      "loss": 5.088,
      "mean_token_accuracy": 0.2175995960831642,
      "num_tokens": 23858185.0,
      "step": 10405
    },
    {
      "entropy": 5.213474702835083,
      "epoch": 1.0,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004908139541891505,
      "loss": 4.8595,
      "mean_token_accuracy": 0.23244536966085433,
      "num_tokens": 23868536.0,
      "step": 10410
    },
    {
      "entropy": 5.313180303573608,
      "epoch": 1.0004803073967339,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004908042556236066,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22595395892858505,
      "num_tokens": 23880283.0,
      "step": 10415
    },
    {
      "entropy": 5.35840950012207,
      "epoch": 1.0009606147934678,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004907945520477286,
      "loss": 4.9749,
      "mean_token_accuracy": 0.22070265114307402,
      "num_tokens": 23892413.0,
      "step": 10420
    },
    {
      "entropy": 5.38995327949524,
      "epoch": 1.0014409221902016,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004907848434617419,
      "loss": 5.0727,
      "mean_token_accuracy": 0.21011357009410858,
      "num_tokens": 23903977.0,
      "step": 10425
    },
    {
      "entropy": 5.39252119064331,
      "epoch": 1.0019212295869357,
      "grad_norm": 0.921875,
      "learning_rate": 0.000490775129865872,
      "loss": 5.0365,
      "mean_token_accuracy": 0.21529491096735,
      "num_tokens": 23915153.0,
      "step": 10430
    },
    {
      "entropy": 5.299467468261719,
      "epoch": 1.0024015369836696,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004907654112603442,
      "loss": 4.9237,
      "mean_token_accuracy": 0.21787198334932328,
      "num_tokens": 23926043.0,
      "step": 10435
    },
    {
      "entropy": 5.209107732772827,
      "epoch": 1.0028818443804035,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004907556876453843,
      "loss": 4.826,
      "mean_token_accuracy": 0.2308427080512047,
      "num_tokens": 23936658.0,
      "step": 10440
    },
    {
      "entropy": 5.2857404232025145,
      "epoch": 1.0033621517771374,
      "grad_norm": 0.9375,
      "learning_rate": 0.000490745959021218,
      "loss": 4.9409,
      "mean_token_accuracy": 0.2155250146985054,
      "num_tokens": 23947676.0,
      "step": 10445
    },
    {
      "entropy": 5.3143213272094725,
      "epoch": 1.0038424591738713,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004907362253880711,
      "loss": 4.9205,
      "mean_token_accuracy": 0.22141096889972686,
      "num_tokens": 23959130.0,
      "step": 10450
    },
    {
      "entropy": 5.380353116989136,
      "epoch": 1.0043227665706052,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004907264867461697,
      "loss": 5.0376,
      "mean_token_accuracy": 0.2110619753599167,
      "num_tokens": 23969905.0,
      "step": 10455
    },
    {
      "entropy": 5.235244274139404,
      "epoch": 1.004803073967339,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004907167430957399,
      "loss": 4.8994,
      "mean_token_accuracy": 0.22810325920581817,
      "num_tokens": 23982016.0,
      "step": 10460
    },
    {
      "entropy": 5.2941690444946286,
      "epoch": 1.005283381364073,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004907069944370077,
      "loss": 5.0465,
      "mean_token_accuracy": 0.21965305060148238,
      "num_tokens": 23994200.0,
      "step": 10465
    },
    {
      "entropy": 5.407249164581299,
      "epoch": 1.005763688760807,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004906972407701998,
      "loss": 4.9385,
      "mean_token_accuracy": 0.23240635693073272,
      "num_tokens": 24004695.0,
      "step": 10470
    },
    {
      "entropy": 5.279523420333862,
      "epoch": 1.006243996157541,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004906874820955423,
      "loss": 4.8751,
      "mean_token_accuracy": 0.2285669967532158,
      "num_tokens": 24015922.0,
      "step": 10475
    },
    {
      "entropy": 5.215548419952393,
      "epoch": 1.0067243035542748,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004906777184132621,
      "loss": 4.8948,
      "mean_token_accuracy": 0.22257906049489976,
      "num_tokens": 24026759.0,
      "step": 10480
    },
    {
      "entropy": 5.306636667251587,
      "epoch": 1.0072046109510087,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004906679497235856,
      "loss": 4.9831,
      "mean_token_accuracy": 0.2136070743203163,
      "num_tokens": 24037988.0,
      "step": 10485
    },
    {
      "entropy": 5.332887840270996,
      "epoch": 1.0076849183477425,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004906581760267397,
      "loss": 4.9988,
      "mean_token_accuracy": 0.2249949038028717,
      "num_tokens": 24050837.0,
      "step": 10490
    },
    {
      "entropy": 5.31575722694397,
      "epoch": 1.0081652257444764,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004906483973229513,
      "loss": 4.8868,
      "mean_token_accuracy": 0.2262268215417862,
      "num_tokens": 24063085.0,
      "step": 10495
    },
    {
      "entropy": 5.262937688827515,
      "epoch": 1.0086455331412103,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004906386136124476,
      "loss": 4.9453,
      "mean_token_accuracy": 0.21875492334365845,
      "num_tokens": 24074082.0,
      "step": 10500
    },
    {
      "entropy": 5.340405654907227,
      "epoch": 1.0091258405379442,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004906288248954554,
      "loss": 4.9832,
      "mean_token_accuracy": 0.22106757909059524,
      "num_tokens": 24085050.0,
      "step": 10505
    },
    {
      "entropy": 5.293986892700195,
      "epoch": 1.0096061479346783,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004906190311722023,
      "loss": 4.9626,
      "mean_token_accuracy": 0.22608766257762908,
      "num_tokens": 24095523.0,
      "step": 10510
    },
    {
      "entropy": 5.225510311126709,
      "epoch": 1.0100864553314122,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004906092324429155,
      "loss": 4.9274,
      "mean_token_accuracy": 0.21385540664196015,
      "num_tokens": 24107157.0,
      "step": 10515
    },
    {
      "entropy": 5.339137601852417,
      "epoch": 1.010566762728146,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004905994287078227,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2251956969499588,
      "num_tokens": 24118668.0,
      "step": 10520
    },
    {
      "entropy": 5.300364828109741,
      "epoch": 1.01104707012488,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004905896199671512,
      "loss": 4.9813,
      "mean_token_accuracy": 0.2235588476061821,
      "num_tokens": 24129563.0,
      "step": 10525
    },
    {
      "entropy": 5.394262504577637,
      "epoch": 1.0115273775216138,
      "grad_norm": 1.0625,
      "learning_rate": 0.000490579806221129,
      "loss": 4.9908,
      "mean_token_accuracy": 0.20981571227312087,
      "num_tokens": 24139965.0,
      "step": 10530
    },
    {
      "entropy": 5.370422601699829,
      "epoch": 1.0120076849183477,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004905699874699838,
      "loss": 4.981,
      "mean_token_accuracy": 0.21888091415166855,
      "num_tokens": 24152109.0,
      "step": 10535
    },
    {
      "entropy": 5.260008096694946,
      "epoch": 1.0124879923150816,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004905601637139436,
      "loss": 4.9521,
      "mean_token_accuracy": 0.2161744639277458,
      "num_tokens": 24162558.0,
      "step": 10540
    },
    {
      "entropy": 5.232922458648682,
      "epoch": 1.0129682997118155,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004905503349532365,
      "loss": 4.8465,
      "mean_token_accuracy": 0.22617940306663514,
      "num_tokens": 24173963.0,
      "step": 10545
    },
    {
      "entropy": 5.269593191146851,
      "epoch": 1.0134486071085496,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004905405011880906,
      "loss": 4.8861,
      "mean_token_accuracy": 0.22899677604436874,
      "num_tokens": 24184369.0,
      "step": 10550
    },
    {
      "entropy": 5.36140079498291,
      "epoch": 1.0139289145052834,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004905306624187343,
      "loss": 5.046,
      "mean_token_accuracy": 0.21708938479423523,
      "num_tokens": 24194692.0,
      "step": 10555
    },
    {
      "entropy": 5.332235193252563,
      "epoch": 1.0144092219020173,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004905208186453961,
      "loss": 4.9674,
      "mean_token_accuracy": 0.21964878290891648,
      "num_tokens": 24205363.0,
      "step": 10560
    },
    {
      "entropy": 5.336849021911621,
      "epoch": 1.0148895292987512,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004905109698683044,
      "loss": 4.9838,
      "mean_token_accuracy": 0.21331555098295213,
      "num_tokens": 24216678.0,
      "step": 10565
    },
    {
      "entropy": 5.23578953742981,
      "epoch": 1.015369836695485,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004905011160876878,
      "loss": 4.873,
      "mean_token_accuracy": 0.21931922435760498,
      "num_tokens": 24227541.0,
      "step": 10570
    },
    {
      "entropy": 5.29215407371521,
      "epoch": 1.015850144092219,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004904912573037753,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22427539378404618,
      "num_tokens": 24238118.0,
      "step": 10575
    },
    {
      "entropy": 5.334722089767456,
      "epoch": 1.0163304514889528,
      "grad_norm": 0.875,
      "learning_rate": 0.0004904813935167957,
      "loss": 5.0111,
      "mean_token_accuracy": 0.21395374536514283,
      "num_tokens": 24250044.0,
      "step": 10580
    },
    {
      "entropy": 5.322008848190308,
      "epoch": 1.0168107588856867,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004904715247269779,
      "loss": 5.022,
      "mean_token_accuracy": 0.2088482603430748,
      "num_tokens": 24262805.0,
      "step": 10585
    },
    {
      "entropy": 5.312357378005982,
      "epoch": 1.0172910662824208,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004904616509345514,
      "loss": 4.9536,
      "mean_token_accuracy": 0.21250459551811218,
      "num_tokens": 24274287.0,
      "step": 10590
    },
    {
      "entropy": 5.183726167678833,
      "epoch": 1.0177713736791547,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004904517721397449,
      "loss": 4.8345,
      "mean_token_accuracy": 0.22320182770490646,
      "num_tokens": 24284839.0,
      "step": 10595
    },
    {
      "entropy": 5.230026292800903,
      "epoch": 1.0182516810758886,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004904418883427881,
      "loss": 4.9465,
      "mean_token_accuracy": 0.22179240137338638,
      "num_tokens": 24295812.0,
      "step": 10600
    },
    {
      "entropy": 5.3224975109100345,
      "epoch": 1.0187319884726225,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004904319995439104,
      "loss": 4.9164,
      "mean_token_accuracy": 0.22933993190526963,
      "num_tokens": 24306365.0,
      "step": 10605
    },
    {
      "entropy": 5.279590511322022,
      "epoch": 1.0192122958693564,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004904221057433412,
      "loss": 4.9151,
      "mean_token_accuracy": 0.2190833196043968,
      "num_tokens": 24317557.0,
      "step": 10610
    },
    {
      "entropy": 5.245771360397339,
      "epoch": 1.0196926032660902,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004904122069413105,
      "loss": 4.8717,
      "mean_token_accuracy": 0.2283520743250847,
      "num_tokens": 24328874.0,
      "step": 10615
    },
    {
      "entropy": 5.384447002410889,
      "epoch": 1.0201729106628241,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000490402303138048,
      "loss": 5.0593,
      "mean_token_accuracy": 0.2126762628555298,
      "num_tokens": 24340918.0,
      "step": 10620
    },
    {
      "entropy": 5.354063701629639,
      "epoch": 1.0206532180595582,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004903923943337836,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21663032472133636,
      "num_tokens": 24352642.0,
      "step": 10625
    },
    {
      "entropy": 5.241005325317383,
      "epoch": 1.021133525456292,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004903824805287475,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2252449706196785,
      "num_tokens": 24364874.0,
      "step": 10630
    },
    {
      "entropy": 5.232929277420044,
      "epoch": 1.021613832853026,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004903725617231696,
      "loss": 4.9782,
      "mean_token_accuracy": 0.22302042543888093,
      "num_tokens": 24376566.0,
      "step": 10635
    },
    {
      "entropy": 5.272281980514526,
      "epoch": 1.0220941402497599,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004903626379172805,
      "loss": 4.8197,
      "mean_token_accuracy": 0.23591640293598176,
      "num_tokens": 24387283.0,
      "step": 10640
    },
    {
      "entropy": 5.365268230438232,
      "epoch": 1.0225744476464937,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004903527091113102,
      "loss": 5.0131,
      "mean_token_accuracy": 0.2176625281572342,
      "num_tokens": 24397970.0,
      "step": 10645
    },
    {
      "entropy": 5.270583534240723,
      "epoch": 1.0230547550432276,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004903427753054897,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21789584457874298,
      "num_tokens": 24411223.0,
      "step": 10650
    },
    {
      "entropy": 5.315692186355591,
      "epoch": 1.0235350624399615,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004903328365000492,
      "loss": 4.9816,
      "mean_token_accuracy": 0.21755286157131196,
      "num_tokens": 24422117.0,
      "step": 10655
    },
    {
      "entropy": 5.283857250213623,
      "epoch": 1.0240153698366954,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004903228926952199,
      "loss": 4.9358,
      "mean_token_accuracy": 0.2208492115139961,
      "num_tokens": 24434142.0,
      "step": 10660
    },
    {
      "entropy": 5.282527732849121,
      "epoch": 1.0244956772334295,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004903129438912322,
      "loss": 4.8399,
      "mean_token_accuracy": 0.2275714933872223,
      "num_tokens": 24445295.0,
      "step": 10665
    },
    {
      "entropy": 5.243832397460937,
      "epoch": 1.0249759846301634,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004903029900883174,
      "loss": 4.9116,
      "mean_token_accuracy": 0.22679072320461274,
      "num_tokens": 24456092.0,
      "step": 10670
    },
    {
      "entropy": 5.3032959461212155,
      "epoch": 1.0254562920268973,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004902930312867063,
      "loss": 5.0205,
      "mean_token_accuracy": 0.2134276196360588,
      "num_tokens": 24467653.0,
      "step": 10675
    },
    {
      "entropy": 5.335419321060181,
      "epoch": 1.0259365994236311,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004902830674866306,
      "loss": 4.9712,
      "mean_token_accuracy": 0.21949176043272017,
      "num_tokens": 24479164.0,
      "step": 10680
    },
    {
      "entropy": 5.230328512191773,
      "epoch": 1.026416906820365,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004902730986883211,
      "loss": 4.8456,
      "mean_token_accuracy": 0.23046619147062303,
      "num_tokens": 24489785.0,
      "step": 10685
    },
    {
      "entropy": 5.289320659637451,
      "epoch": 1.026897214217099,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004902631248920096,
      "loss": 4.9382,
      "mean_token_accuracy": 0.2251667246222496,
      "num_tokens": 24500158.0,
      "step": 10690
    },
    {
      "entropy": 5.340549182891846,
      "epoch": 1.0273775216138328,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004902531460979274,
      "loss": 5.0493,
      "mean_token_accuracy": 0.2103318139910698,
      "num_tokens": 24512851.0,
      "step": 10695
    },
    {
      "entropy": 5.315286016464233,
      "epoch": 1.0278578290105667,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004902431623063065,
      "loss": 4.9889,
      "mean_token_accuracy": 0.22032700330018998,
      "num_tokens": 24524016.0,
      "step": 10700
    },
    {
      "entropy": 5.203576374053955,
      "epoch": 1.0283381364073008,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004902331735173785,
      "loss": 4.8701,
      "mean_token_accuracy": 0.21959290653467178,
      "num_tokens": 24536348.0,
      "step": 10705
    },
    {
      "entropy": 5.228856945037842,
      "epoch": 1.0288184438040346,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004902231797313752,
      "loss": 4.9382,
      "mean_token_accuracy": 0.21476825475692748,
      "num_tokens": 24548718.0,
      "step": 10710
    },
    {
      "entropy": 5.243882083892823,
      "epoch": 1.0292987512007685,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004902131809485288,
      "loss": 4.8457,
      "mean_token_accuracy": 0.2360962972044945,
      "num_tokens": 24560567.0,
      "step": 10715
    },
    {
      "entropy": 5.259013319015503,
      "epoch": 1.0297790585975024,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004902031771690713,
      "loss": 4.8795,
      "mean_token_accuracy": 0.22278717011213303,
      "num_tokens": 24572610.0,
      "step": 10720
    },
    {
      "entropy": 5.285334920883178,
      "epoch": 1.0302593659942363,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004901931683932352,
      "loss": 4.917,
      "mean_token_accuracy": 0.22453642934560775,
      "num_tokens": 24584738.0,
      "step": 10725
    },
    {
      "entropy": 5.321272563934326,
      "epoch": 1.0307396733909702,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004901831546212526,
      "loss": 4.9658,
      "mean_token_accuracy": 0.22291696667671204,
      "num_tokens": 24596603.0,
      "step": 10730
    },
    {
      "entropy": 5.258141136169433,
      "epoch": 1.031219980787704,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004901731358533562,
      "loss": 4.8592,
      "mean_token_accuracy": 0.2240314483642578,
      "num_tokens": 24607061.0,
      "step": 10735
    },
    {
      "entropy": 5.211468696594238,
      "epoch": 1.031700288184438,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004901631120897785,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22234783917665482,
      "num_tokens": 24619177.0,
      "step": 10740
    },
    {
      "entropy": 5.216166257858276,
      "epoch": 1.032180595581172,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004901530833307522,
      "loss": 4.8524,
      "mean_token_accuracy": 0.22541184574365616,
      "num_tokens": 24631336.0,
      "step": 10745
    },
    {
      "entropy": 5.280277347564697,
      "epoch": 1.032660902977906,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004901430495765103,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2319018304347992,
      "num_tokens": 24641743.0,
      "step": 10750
    },
    {
      "entropy": 5.250428533554077,
      "epoch": 1.0331412103746398,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004901330108272855,
      "loss": 4.8639,
      "mean_token_accuracy": 0.22607083916664122,
      "num_tokens": 24652318.0,
      "step": 10755
    },
    {
      "entropy": 5.25404920578003,
      "epoch": 1.0336215177713737,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004901229670833111,
      "loss": 4.8882,
      "mean_token_accuracy": 0.22083948403596879,
      "num_tokens": 24664129.0,
      "step": 10760
    },
    {
      "entropy": 5.335971355438232,
      "epoch": 1.0341018251681076,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004901129183448201,
      "loss": 4.9614,
      "mean_token_accuracy": 0.21500274240970613,
      "num_tokens": 24674921.0,
      "step": 10765
    },
    {
      "entropy": 5.2650950908660885,
      "epoch": 1.0345821325648414,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004901028646120459,
      "loss": 4.9228,
      "mean_token_accuracy": 0.21792251914739608,
      "num_tokens": 24686052.0,
      "step": 10770
    },
    {
      "entropy": 5.248932123184204,
      "epoch": 1.0350624399615753,
      "grad_norm": 0.828125,
      "learning_rate": 0.000490092805885222,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22360815703868867,
      "num_tokens": 24698625.0,
      "step": 10775
    },
    {
      "entropy": 5.249554061889649,
      "epoch": 1.0355427473583094,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004900827421645816,
      "loss": 4.8743,
      "mean_token_accuracy": 0.23170357197523117,
      "num_tokens": 24709322.0,
      "step": 10780
    },
    {
      "entropy": 5.2204649448394775,
      "epoch": 1.0360230547550433,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004900726734503589,
      "loss": 4.8538,
      "mean_token_accuracy": 0.22541843354701996,
      "num_tokens": 24719512.0,
      "step": 10785
    },
    {
      "entropy": 5.244796228408814,
      "epoch": 1.0365033621517772,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004900625997427872,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2267092987895012,
      "num_tokens": 24729947.0,
      "step": 10790
    },
    {
      "entropy": 5.183066749572754,
      "epoch": 1.036983669548511,
      "grad_norm": 1.0,
      "learning_rate": 0.0004900525210421006,
      "loss": 4.875,
      "mean_token_accuracy": 0.2356373742222786,
      "num_tokens": 24741423.0,
      "step": 10795
    },
    {
      "entropy": 5.275487327575684,
      "epoch": 1.037463976945245,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004900424373485329,
      "loss": 4.9188,
      "mean_token_accuracy": 0.22111906111240387,
      "num_tokens": 24753403.0,
      "step": 10800
    },
    {
      "entropy": 5.339709854125976,
      "epoch": 1.0379442843419788,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004900323486623185,
      "loss": 5.0244,
      "mean_token_accuracy": 0.21418058574199678,
      "num_tokens": 24763660.0,
      "step": 10805
    },
    {
      "entropy": 5.28852276802063,
      "epoch": 1.0384245917387127,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004900222549836914,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22718273103237152,
      "num_tokens": 24775061.0,
      "step": 10810
    },
    {
      "entropy": 5.2988893508911135,
      "epoch": 1.0389048991354466,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000490012156312886,
      "loss": 4.9289,
      "mean_token_accuracy": 0.21546023041009904,
      "num_tokens": 24785248.0,
      "step": 10815
    },
    {
      "entropy": 5.214194250106812,
      "epoch": 1.0393852065321807,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004900020526501369,
      "loss": 4.8228,
      "mean_token_accuracy": 0.2314630925655365,
      "num_tokens": 24797024.0,
      "step": 10820
    },
    {
      "entropy": 5.38393006324768,
      "epoch": 1.0398655139289146,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004899919439956785,
      "loss": 5.0762,
      "mean_token_accuracy": 0.21064777225255965,
      "num_tokens": 24808500.0,
      "step": 10825
    },
    {
      "entropy": 5.378734493255616,
      "epoch": 1.0403458213256485,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004899818303497455,
      "loss": 5.0267,
      "mean_token_accuracy": 0.2148999571800232,
      "num_tokens": 24818805.0,
      "step": 10830
    },
    {
      "entropy": 5.318168210983276,
      "epoch": 1.0408261287223823,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004899717117125728,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21729080379009247,
      "num_tokens": 24829247.0,
      "step": 10835
    },
    {
      "entropy": 5.207594299316407,
      "epoch": 1.0413064361191162,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004899615880843953,
      "loss": 4.8849,
      "mean_token_accuracy": 0.22602875977754594,
      "num_tokens": 24840139.0,
      "step": 10840
    },
    {
      "entropy": 5.320698928833008,
      "epoch": 1.04178674351585,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004899514594654481,
      "loss": 4.989,
      "mean_token_accuracy": 0.21491976082324982,
      "num_tokens": 24851734.0,
      "step": 10845
    },
    {
      "entropy": 5.296469736099243,
      "epoch": 1.042267050912584,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004899413258559662,
      "loss": 4.9528,
      "mean_token_accuracy": 0.22285531163215638,
      "num_tokens": 24863424.0,
      "step": 10850
    },
    {
      "entropy": 5.2722986221313475,
      "epoch": 1.0427473583093179,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004899311872561849,
      "loss": 4.967,
      "mean_token_accuracy": 0.21983652859926223,
      "num_tokens": 24875086.0,
      "step": 10855
    },
    {
      "entropy": 5.267477655410767,
      "epoch": 1.043227665706052,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004899210436663398,
      "loss": 4.9008,
      "mean_token_accuracy": 0.2237691879272461,
      "num_tokens": 24888408.0,
      "step": 10860
    },
    {
      "entropy": 5.23045506477356,
      "epoch": 1.0437079731027858,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004899108950866661,
      "loss": 4.8367,
      "mean_token_accuracy": 0.23072488307952882,
      "num_tokens": 24900357.0,
      "step": 10865
    },
    {
      "entropy": 5.275755786895752,
      "epoch": 1.0441882804995197,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004899007415173997,
      "loss": 4.9442,
      "mean_token_accuracy": 0.22494025230407716,
      "num_tokens": 24910790.0,
      "step": 10870
    },
    {
      "entropy": 5.268125009536743,
      "epoch": 1.0446685878962536,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004898905829587762,
      "loss": 4.9212,
      "mean_token_accuracy": 0.22150873392820358,
      "num_tokens": 24922124.0,
      "step": 10875
    },
    {
      "entropy": 5.196393775939941,
      "epoch": 1.0451488952929875,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004898804194110313,
      "loss": 4.8246,
      "mean_token_accuracy": 0.2372725561261177,
      "num_tokens": 24933591.0,
      "step": 10880
    },
    {
      "entropy": 5.235292959213257,
      "epoch": 1.0456292026897214,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004898702508744012,
      "loss": 4.8488,
      "mean_token_accuracy": 0.22452994734048842,
      "num_tokens": 24944708.0,
      "step": 10885
    },
    {
      "entropy": 5.224345541000366,
      "epoch": 1.0461095100864553,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004898600773491221,
      "loss": 4.8235,
      "mean_token_accuracy": 0.2307368054986,
      "num_tokens": 24955966.0,
      "step": 10890
    },
    {
      "entropy": 5.30643949508667,
      "epoch": 1.0465898174831891,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004898498988354297,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22580115497112274,
      "num_tokens": 24967292.0,
      "step": 10895
    },
    {
      "entropy": 5.294908809661865,
      "epoch": 1.0470701248799232,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004898397153335608,
      "loss": 4.9827,
      "mean_token_accuracy": 0.21854006201028825,
      "num_tokens": 24977407.0,
      "step": 10900
    },
    {
      "entropy": 5.366672945022583,
      "epoch": 1.0475504322766571,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004898295268437517,
      "loss": 5.0347,
      "mean_token_accuracy": 0.22023238092660904,
      "num_tokens": 24988804.0,
      "step": 10905
    },
    {
      "entropy": 5.297868251800537,
      "epoch": 1.048030739673391,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004898193333662388,
      "loss": 4.9808,
      "mean_token_accuracy": 0.21805969029664993,
      "num_tokens": 25000297.0,
      "step": 10910
    },
    {
      "entropy": 5.283839416503906,
      "epoch": 1.0485110470701249,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004898091349012588,
      "loss": 4.9032,
      "mean_token_accuracy": 0.22473251968622207,
      "num_tokens": 25012135.0,
      "step": 10915
    },
    {
      "entropy": 5.2092221736907955,
      "epoch": 1.0489913544668588,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004897989314490486,
      "loss": 4.891,
      "mean_token_accuracy": 0.23073551952838897,
      "num_tokens": 25023572.0,
      "step": 10920
    },
    {
      "entropy": 5.285353374481201,
      "epoch": 1.0494716618635926,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004897887230098451,
      "loss": 4.943,
      "mean_token_accuracy": 0.22414311319589614,
      "num_tokens": 25035015.0,
      "step": 10925
    },
    {
      "entropy": 5.276854658126831,
      "epoch": 1.0499519692603265,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004897785095838852,
      "loss": 4.8982,
      "mean_token_accuracy": 0.21893301457166672,
      "num_tokens": 25045931.0,
      "step": 10930
    },
    {
      "entropy": 5.3147846221923825,
      "epoch": 1.0504322766570606,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004897682911714061,
      "loss": 4.9317,
      "mean_token_accuracy": 0.2254469946026802,
      "num_tokens": 25056767.0,
      "step": 10935
    },
    {
      "entropy": 5.217756128311157,
      "epoch": 1.0509125840537945,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000489758067772645,
      "loss": 4.886,
      "mean_token_accuracy": 0.2318438395857811,
      "num_tokens": 25068731.0,
      "step": 10940
    },
    {
      "entropy": 5.250807046890259,
      "epoch": 1.0513928914505284,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004897478393878392,
      "loss": 4.9655,
      "mean_token_accuracy": 0.21776752471923827,
      "num_tokens": 25081268.0,
      "step": 10945
    },
    {
      "entropy": 5.210177326202393,
      "epoch": 1.0518731988472623,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004897376060172264,
      "loss": 4.872,
      "mean_token_accuracy": 0.2283642113208771,
      "num_tokens": 25093105.0,
      "step": 10950
    },
    {
      "entropy": 5.253404569625855,
      "epoch": 1.0523535062439962,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004897273676610438,
      "loss": 4.8693,
      "mean_token_accuracy": 0.22874182909727098,
      "num_tokens": 25103766.0,
      "step": 10955
    },
    {
      "entropy": 5.303346395492554,
      "epoch": 1.05283381364073,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004897171243195295,
      "loss": 5.0127,
      "mean_token_accuracy": 0.2204292520880699,
      "num_tokens": 25115675.0,
      "step": 10960
    },
    {
      "entropy": 5.270561599731446,
      "epoch": 1.053314121037464,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000489706875992921,
      "loss": 4.8784,
      "mean_token_accuracy": 0.22554850429296494,
      "num_tokens": 25127907.0,
      "step": 10965
    },
    {
      "entropy": 5.2606603622436525,
      "epoch": 1.0537944284341978,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004896966226814565,
      "loss": 4.916,
      "mean_token_accuracy": 0.22626586705446244,
      "num_tokens": 25139675.0,
      "step": 10970
    },
    {
      "entropy": 5.332069349288941,
      "epoch": 1.054274735830932,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004896863643853739,
      "loss": 5.0215,
      "mean_token_accuracy": 0.22300835996866225,
      "num_tokens": 25150960.0,
      "step": 10975
    },
    {
      "entropy": 5.316552019119262,
      "epoch": 1.0547550432276658,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896761011049114,
      "loss": 4.9945,
      "mean_token_accuracy": 0.2180240347981453,
      "num_tokens": 25163676.0,
      "step": 10980
    },
    {
      "entropy": 5.230193948745727,
      "epoch": 1.0552353506243997,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004896658328403074,
      "loss": 4.9325,
      "mean_token_accuracy": 0.21965030580759048,
      "num_tokens": 25174317.0,
      "step": 10985
    },
    {
      "entropy": 5.2390213966369625,
      "epoch": 1.0557156580211335,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004896555595918001,
      "loss": 4.9327,
      "mean_token_accuracy": 0.220357546210289,
      "num_tokens": 25186585.0,
      "step": 10990
    },
    {
      "entropy": 5.292184495925904,
      "epoch": 1.0561959654178674,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004896452813596281,
      "loss": 4.9532,
      "mean_token_accuracy": 0.21673746258020402,
      "num_tokens": 25199014.0,
      "step": 10995
    },
    {
      "entropy": 5.344906234741211,
      "epoch": 1.0566762728146013,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896349981440301,
      "loss": 5.049,
      "mean_token_accuracy": 0.20951563268899917,
      "num_tokens": 25210544.0,
      "step": 11000
    },
    {
      "entropy": 5.292508840560913,
      "epoch": 1.0571565802113352,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004896247099452447,
      "loss": 4.9353,
      "mean_token_accuracy": 0.22783829122781754,
      "num_tokens": 25221583.0,
      "step": 11005
    },
    {
      "entropy": 5.293026733398437,
      "epoch": 1.057636887608069,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004896144167635108,
      "loss": 4.9218,
      "mean_token_accuracy": 0.22860798835754395,
      "num_tokens": 25231724.0,
      "step": 11010
    },
    {
      "entropy": 5.30416898727417,
      "epoch": 1.0581171950048032,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004896041185990675,
      "loss": 5.009,
      "mean_token_accuracy": 0.2183375522494316,
      "num_tokens": 25243021.0,
      "step": 11015
    },
    {
      "entropy": 5.337156057357788,
      "epoch": 1.058597502401537,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004895938154521538,
      "loss": 4.9905,
      "mean_token_accuracy": 0.2188921794295311,
      "num_tokens": 25254189.0,
      "step": 11020
    },
    {
      "entropy": 5.3018392562866214,
      "epoch": 1.059077809798271,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004895835073230089,
      "loss": 5.0084,
      "mean_token_accuracy": 0.22265112847089769,
      "num_tokens": 25265556.0,
      "step": 11025
    },
    {
      "entropy": 5.208692026138306,
      "epoch": 1.0595581171950048,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004895731942118722,
      "loss": 4.8512,
      "mean_token_accuracy": 0.22401011735200882,
      "num_tokens": 25276789.0,
      "step": 11030
    },
    {
      "entropy": 5.297592687606811,
      "epoch": 1.0600384245917387,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004895628761189829,
      "loss": 5.0229,
      "mean_token_accuracy": 0.21535513252019883,
      "num_tokens": 25288505.0,
      "step": 11035
    },
    {
      "entropy": 5.243595457077026,
      "epoch": 1.0605187319884726,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004895525530445809,
      "loss": 4.8646,
      "mean_token_accuracy": 0.23527745008468628,
      "num_tokens": 25301490.0,
      "step": 11040
    },
    {
      "entropy": 5.279550552368164,
      "epoch": 1.0609990393852065,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004895422249889057,
      "loss": 4.9604,
      "mean_token_accuracy": 0.21967684030532836,
      "num_tokens": 25313303.0,
      "step": 11045
    },
    {
      "entropy": 5.38269739151001,
      "epoch": 1.0614793467819403,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004895318919521971,
      "loss": 5.019,
      "mean_token_accuracy": 0.209059377014637,
      "num_tokens": 25324379.0,
      "step": 11050
    },
    {
      "entropy": 5.2174498558044435,
      "epoch": 1.0619596541786744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004895215539346949,
      "loss": 4.857,
      "mean_token_accuracy": 0.22579782903194429,
      "num_tokens": 25335834.0,
      "step": 11055
    },
    {
      "entropy": 5.252145862579345,
      "epoch": 1.0624399615754083,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004895112109366393,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2212860882282257,
      "num_tokens": 25347591.0,
      "step": 11060
    },
    {
      "entropy": 5.2734057903289795,
      "epoch": 1.0629202689721422,
      "grad_norm": 1.0,
      "learning_rate": 0.0004895008629582703,
      "loss": 4.912,
      "mean_token_accuracy": 0.22627960294485092,
      "num_tokens": 25358483.0,
      "step": 11065
    },
    {
      "entropy": 5.252391195297241,
      "epoch": 1.063400576368876,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004894905099998283,
      "loss": 4.9233,
      "mean_token_accuracy": 0.22080318480730057,
      "num_tokens": 25369434.0,
      "step": 11070
    },
    {
      "entropy": 5.259957885742187,
      "epoch": 1.06388088376561,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004894801520615535,
      "loss": 4.9321,
      "mean_token_accuracy": 0.22465706914663314,
      "num_tokens": 25381007.0,
      "step": 11075
    },
    {
      "entropy": 5.252924585342408,
      "epoch": 1.0643611911623438,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004894697891436863,
      "loss": 4.9151,
      "mean_token_accuracy": 0.21939058601856232,
      "num_tokens": 25393809.0,
      "step": 11080
    },
    {
      "entropy": 5.31340537071228,
      "epoch": 1.0648414985590777,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004894594212464676,
      "loss": 4.9542,
      "mean_token_accuracy": 0.22666523456573487,
      "num_tokens": 25404967.0,
      "step": 11085
    },
    {
      "entropy": 5.286955213546753,
      "epoch": 1.0653218059558118,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004894490483701381,
      "loss": 4.9055,
      "mean_token_accuracy": 0.23343347311019896,
      "num_tokens": 25417092.0,
      "step": 11090
    },
    {
      "entropy": 5.308970880508423,
      "epoch": 1.0658021133525457,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004894386705149382,
      "loss": 4.9892,
      "mean_token_accuracy": 0.21542260497808458,
      "num_tokens": 25428425.0,
      "step": 11095
    },
    {
      "entropy": 5.258173131942749,
      "epoch": 1.0662824207492796,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004894282876811093,
      "loss": 4.8245,
      "mean_token_accuracy": 0.2359119400382042,
      "num_tokens": 25440134.0,
      "step": 11100
    },
    {
      "entropy": 5.209200143814087,
      "epoch": 1.0667627281460135,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004894178998688921,
      "loss": 4.9042,
      "mean_token_accuracy": 0.22675962895154952,
      "num_tokens": 25452222.0,
      "step": 11105
    },
    {
      "entropy": 5.242311620712281,
      "epoch": 1.0672430355427474,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004894075070785281,
      "loss": 4.941,
      "mean_token_accuracy": 0.22396333366632462,
      "num_tokens": 25464541.0,
      "step": 11110
    },
    {
      "entropy": 5.3597979068756105,
      "epoch": 1.0677233429394812,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004893971093102585,
      "loss": 4.954,
      "mean_token_accuracy": 0.21432135999202728,
      "num_tokens": 25476537.0,
      "step": 11115
    },
    {
      "entropy": 5.393908405303955,
      "epoch": 1.0682036503362151,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004893867065643245,
      "loss": 4.9999,
      "mean_token_accuracy": 0.2121525451540947,
      "num_tokens": 25486737.0,
      "step": 11120
    },
    {
      "entropy": 5.23684401512146,
      "epoch": 1.068683957732949,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004893762988409678,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21824057549238204,
      "num_tokens": 25497278.0,
      "step": 11125
    },
    {
      "entropy": 5.247014665603638,
      "epoch": 1.069164265129683,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004893658861404301,
      "loss": 4.9157,
      "mean_token_accuracy": 0.21818040758371354,
      "num_tokens": 25508716.0,
      "step": 11130
    },
    {
      "entropy": 5.363707494735718,
      "epoch": 1.069644572526417,
      "grad_norm": 1.0,
      "learning_rate": 0.0004893554684629529,
      "loss": 5.0114,
      "mean_token_accuracy": 0.2178993597626686,
      "num_tokens": 25519439.0,
      "step": 11135
    },
    {
      "entropy": 5.239768743515015,
      "epoch": 1.0701248799231509,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004893450458087784,
      "loss": 4.8026,
      "mean_token_accuracy": 0.23096322864294053,
      "num_tokens": 25530911.0,
      "step": 11140
    },
    {
      "entropy": 5.309357023239135,
      "epoch": 1.0706051873198847,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004893346181781483,
      "loss": 5.045,
      "mean_token_accuracy": 0.20911059379577637,
      "num_tokens": 25542452.0,
      "step": 11145
    },
    {
      "entropy": 5.290190887451172,
      "epoch": 1.0710854947166186,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004893241855713048,
      "loss": 4.9508,
      "mean_token_accuracy": 0.2088031843304634,
      "num_tokens": 25554105.0,
      "step": 11150
    },
    {
      "entropy": 5.3907349586486815,
      "epoch": 1.0715658021133525,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004893137479884903,
      "loss": 5.0139,
      "mean_token_accuracy": 0.21952938437461852,
      "num_tokens": 25564806.0,
      "step": 11155
    },
    {
      "entropy": 5.23767466545105,
      "epoch": 1.0720461095100864,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004893033054299468,
      "loss": 4.9094,
      "mean_token_accuracy": 0.22624558806419373,
      "num_tokens": 25575664.0,
      "step": 11160
    },
    {
      "entropy": 5.235759353637695,
      "epoch": 1.0725264169068203,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000489292857895917,
      "loss": 4.8283,
      "mean_token_accuracy": 0.23185605704784393,
      "num_tokens": 25586227.0,
      "step": 11165
    },
    {
      "entropy": 5.30611834526062,
      "epoch": 1.0730067243035544,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004892824053866432,
      "loss": 4.9989,
      "mean_token_accuracy": 0.2191235229372978,
      "num_tokens": 25597475.0,
      "step": 11170
    },
    {
      "entropy": 5.235492897033692,
      "epoch": 1.0734870317002883,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004892719479023683,
      "loss": 4.8883,
      "mean_token_accuracy": 0.2306048735976219,
      "num_tokens": 25608098.0,
      "step": 11175
    },
    {
      "entropy": 5.318383264541626,
      "epoch": 1.0739673390970221,
      "grad_norm": 0.953125,
      "learning_rate": 0.000489261485443335,
      "loss": 4.9911,
      "mean_token_accuracy": 0.218204003572464,
      "num_tokens": 25620053.0,
      "step": 11180
    },
    {
      "entropy": 5.325300598144532,
      "epoch": 1.074447646493756,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004892510180097863,
      "loss": 4.9243,
      "mean_token_accuracy": 0.2171517416834831,
      "num_tokens": 25630534.0,
      "step": 11185
    },
    {
      "entropy": 5.355429172515869,
      "epoch": 1.07492795389049,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004892405456019651,
      "loss": 4.9494,
      "mean_token_accuracy": 0.22859165519475938,
      "num_tokens": 25641413.0,
      "step": 11190
    },
    {
      "entropy": 5.20083909034729,
      "epoch": 1.0754082612872238,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004892300682201147,
      "loss": 4.883,
      "mean_token_accuracy": 0.2301346093416214,
      "num_tokens": 25652081.0,
      "step": 11195
    },
    {
      "entropy": 5.274270820617676,
      "epoch": 1.0758885686839577,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004892195858644782,
      "loss": 4.9778,
      "mean_token_accuracy": 0.22848468720912934,
      "num_tokens": 25664282.0,
      "step": 11200
    },
    {
      "entropy": 5.3464099884033205,
      "epoch": 1.0763688760806915,
      "grad_norm": 0.953125,
      "learning_rate": 0.000489209098535299,
      "loss": 5.0515,
      "mean_token_accuracy": 0.22116044610738755,
      "num_tokens": 25675310.0,
      "step": 11205
    },
    {
      "entropy": 5.403557348251343,
      "epoch": 1.0768491834774256,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004891986062328205,
      "loss": 5.0265,
      "mean_token_accuracy": 0.21974188834428787,
      "num_tokens": 25686895.0,
      "step": 11210
    },
    {
      "entropy": 5.244028472900391,
      "epoch": 1.0773294908741595,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004891881089572865,
      "loss": 4.827,
      "mean_token_accuracy": 0.22494327127933503,
      "num_tokens": 25697778.0,
      "step": 11215
    },
    {
      "entropy": 5.236339426040649,
      "epoch": 1.0778097982708934,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004891776067089406,
      "loss": 4.9736,
      "mean_token_accuracy": 0.21784978955984116,
      "num_tokens": 25708602.0,
      "step": 11220
    },
    {
      "entropy": 5.2394795417785645,
      "epoch": 1.0782901056676273,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004891670994880266,
      "loss": 4.8604,
      "mean_token_accuracy": 0.23020517975091934,
      "num_tokens": 25719671.0,
      "step": 11225
    },
    {
      "entropy": 5.335706329345703,
      "epoch": 1.0787704130643612,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004891565872947888,
      "loss": 5.0393,
      "mean_token_accuracy": 0.2153046026825905,
      "num_tokens": 25731797.0,
      "step": 11230
    },
    {
      "entropy": 5.25079836845398,
      "epoch": 1.079250720461095,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004891460701294706,
      "loss": 4.9019,
      "mean_token_accuracy": 0.22723452895879745,
      "num_tokens": 25743984.0,
      "step": 11235
    },
    {
      "entropy": 5.240848207473755,
      "epoch": 1.079731027857829,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004891355479923167,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22551220059394836,
      "num_tokens": 25755252.0,
      "step": 11240
    },
    {
      "entropy": 5.200658559799194,
      "epoch": 1.080211335254563,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004891250208835712,
      "loss": 4.8285,
      "mean_token_accuracy": 0.2252291589975357,
      "num_tokens": 25765715.0,
      "step": 11245
    },
    {
      "entropy": 5.325733232498169,
      "epoch": 1.080691642651297,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004891144888034784,
      "loss": 4.9621,
      "mean_token_accuracy": 0.21792520433664322,
      "num_tokens": 25777866.0,
      "step": 11250
    },
    {
      "entropy": 5.270968246459961,
      "epoch": 1.0811719500480308,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004891039517522832,
      "loss": 4.8678,
      "mean_token_accuracy": 0.2267429083585739,
      "num_tokens": 25791199.0,
      "step": 11255
    },
    {
      "entropy": 5.186959457397461,
      "epoch": 1.0816522574447647,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004890934097302299,
      "loss": 4.8234,
      "mean_token_accuracy": 0.23791683614253997,
      "num_tokens": 25802979.0,
      "step": 11260
    },
    {
      "entropy": 5.164671754837036,
      "epoch": 1.0821325648414986,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004890828627375632,
      "loss": 4.8602,
      "mean_token_accuracy": 0.22242806106805801,
      "num_tokens": 25814696.0,
      "step": 11265
    },
    {
      "entropy": 5.307852506637573,
      "epoch": 1.0826128722382324,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004890723107745283,
      "loss": 4.9614,
      "mean_token_accuracy": 0.21362551897764206,
      "num_tokens": 25825376.0,
      "step": 11270
    },
    {
      "entropy": 5.28998818397522,
      "epoch": 1.0830931796349663,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004890617538413699,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22705887109041215,
      "num_tokens": 25835491.0,
      "step": 11275
    },
    {
      "entropy": 5.276081895828247,
      "epoch": 1.0835734870317002,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004890511919383333,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22158263772726058,
      "num_tokens": 25848154.0,
      "step": 11280
    },
    {
      "entropy": 5.263273239135742,
      "epoch": 1.084053794428434,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004890406250656636,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22299245595932007,
      "num_tokens": 25859471.0,
      "step": 11285
    },
    {
      "entropy": 5.220771074295044,
      "epoch": 1.0845341018251682,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004890300532236062,
      "loss": 4.8682,
      "mean_token_accuracy": 0.22894442230463027,
      "num_tokens": 25869460.0,
      "step": 11290
    },
    {
      "entropy": 5.279849576950073,
      "epoch": 1.085014409221902,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004890194764124064,
      "loss": 4.8907,
      "mean_token_accuracy": 0.22437620311975479,
      "num_tokens": 25881892.0,
      "step": 11295
    },
    {
      "entropy": 5.3110363483428955,
      "epoch": 1.085494716618636,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004890088946323099,
      "loss": 5.0119,
      "mean_token_accuracy": 0.2122369810938835,
      "num_tokens": 25893774.0,
      "step": 11300
    },
    {
      "entropy": 5.209962940216064,
      "epoch": 1.0859750240153698,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004889983078835623,
      "loss": 4.8287,
      "mean_token_accuracy": 0.22523492872714995,
      "num_tokens": 25904758.0,
      "step": 11305
    },
    {
      "entropy": 5.2393111228942875,
      "epoch": 1.0864553314121037,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004889877161664096,
      "loss": 4.9472,
      "mean_token_accuracy": 0.22046012580394744,
      "num_tokens": 25917013.0,
      "step": 11310
    },
    {
      "entropy": 5.361552333831787,
      "epoch": 1.0869356388088376,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004889771194810974,
      "loss": 4.9968,
      "mean_token_accuracy": 0.22450659275054932,
      "num_tokens": 25927780.0,
      "step": 11315
    },
    {
      "entropy": 5.233099174499512,
      "epoch": 1.0874159462055715,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004889665178278719,
      "loss": 4.8941,
      "mean_token_accuracy": 0.23087696731090546,
      "num_tokens": 25939339.0,
      "step": 11320
    },
    {
      "entropy": 5.274503135681153,
      "epoch": 1.0878962536023056,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004889559112069792,
      "loss": 4.8814,
      "mean_token_accuracy": 0.22502919137477875,
      "num_tokens": 25950440.0,
      "step": 11325
    },
    {
      "entropy": 5.284966850280762,
      "epoch": 1.0883765609990395,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004889452996186657,
      "loss": 4.9336,
      "mean_token_accuracy": 0.2203456610441208,
      "num_tokens": 25962849.0,
      "step": 11330
    },
    {
      "entropy": 5.284975051879883,
      "epoch": 1.0888568683957733,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004889346830631774,
      "loss": 4.9563,
      "mean_token_accuracy": 0.21937913596630096,
      "num_tokens": 25973616.0,
      "step": 11335
    },
    {
      "entropy": 5.232392263412476,
      "epoch": 1.0893371757925072,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000488924061540761,
      "loss": 4.8476,
      "mean_token_accuracy": 0.23240684866905212,
      "num_tokens": 25984727.0,
      "step": 11340
    },
    {
      "entropy": 5.300498723983765,
      "epoch": 1.089817483189241,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004889134350516633,
      "loss": 4.9611,
      "mean_token_accuracy": 0.22785247713327408,
      "num_tokens": 25996431.0,
      "step": 11345
    },
    {
      "entropy": 5.30584716796875,
      "epoch": 1.090297790585975,
      "grad_norm": 0.875,
      "learning_rate": 0.0004889028035961308,
      "loss": 4.9039,
      "mean_token_accuracy": 0.220453242957592,
      "num_tokens": 26008936.0,
      "step": 11350
    },
    {
      "entropy": 5.346635818481445,
      "epoch": 1.0907780979827089,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004888921671744103,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2125432685017586,
      "num_tokens": 26019308.0,
      "step": 11355
    },
    {
      "entropy": 5.290782737731933,
      "epoch": 1.0912584053794427,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004888815257867488,
      "loss": 4.8789,
      "mean_token_accuracy": 0.23098173290491103,
      "num_tokens": 26030705.0,
      "step": 11360
    },
    {
      "entropy": 5.270463562011718,
      "epoch": 1.0917387127761768,
      "grad_norm": 0.875,
      "learning_rate": 0.0004888708794333934,
      "loss": 4.8989,
      "mean_token_accuracy": 0.2233803391456604,
      "num_tokens": 26042759.0,
      "step": 11365
    },
    {
      "entropy": 5.276461935043335,
      "epoch": 1.0922190201729107,
      "grad_norm": 0.875,
      "learning_rate": 0.0004888602281145913,
      "loss": 4.9664,
      "mean_token_accuracy": 0.22298839688301086,
      "num_tokens": 26054719.0,
      "step": 11370
    },
    {
      "entropy": 5.25107626914978,
      "epoch": 1.0926993275696446,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004888495718305897,
      "loss": 4.9576,
      "mean_token_accuracy": 0.21599542945623398,
      "num_tokens": 26065765.0,
      "step": 11375
    },
    {
      "entropy": 5.145067310333252,
      "epoch": 1.0931796349663785,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000488838910581636,
      "loss": 4.8601,
      "mean_token_accuracy": 0.23068887740373611,
      "num_tokens": 26077719.0,
      "step": 11380
    },
    {
      "entropy": 5.314440155029297,
      "epoch": 1.0936599423631124,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004888282443679777,
      "loss": 4.999,
      "mean_token_accuracy": 0.21205778270959855,
      "num_tokens": 26089924.0,
      "step": 11385
    },
    {
      "entropy": 5.4366716861724855,
      "epoch": 1.0941402497598463,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004888175731898627,
      "loss": 5.0538,
      "mean_token_accuracy": 0.21303649097681046,
      "num_tokens": 26100312.0,
      "step": 11390
    },
    {
      "entropy": 5.272197818756103,
      "epoch": 1.0946205571565801,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004888068970475384,
      "loss": 4.9478,
      "mean_token_accuracy": 0.22318591326475143,
      "num_tokens": 26111932.0,
      "step": 11395
    },
    {
      "entropy": 5.272916603088379,
      "epoch": 1.0951008645533142,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004887962159412529,
      "loss": 4.8567,
      "mean_token_accuracy": 0.22836549282073976,
      "num_tokens": 26123989.0,
      "step": 11400
    },
    {
      "entropy": 5.3013030052185055,
      "epoch": 1.0955811719500481,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004887855298712541,
      "loss": 4.9669,
      "mean_token_accuracy": 0.2208797365427017,
      "num_tokens": 26135589.0,
      "step": 11405
    },
    {
      "entropy": 5.190200901031494,
      "epoch": 1.096061479346782,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048877483883779,
      "loss": 4.8819,
      "mean_token_accuracy": 0.22563115358352662,
      "num_tokens": 26147069.0,
      "step": 11410
    },
    {
      "entropy": 5.269239902496338,
      "epoch": 1.0965417867435159,
      "grad_norm": 0.875,
      "learning_rate": 0.0004887641428411091,
      "loss": 4.9382,
      "mean_token_accuracy": 0.22136942893266678,
      "num_tokens": 26159331.0,
      "step": 11415
    },
    {
      "entropy": 5.288673973083496,
      "epoch": 1.0970220941402498,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004887534418814595,
      "loss": 4.9321,
      "mean_token_accuracy": 0.22562492936849593,
      "num_tokens": 26169863.0,
      "step": 11420
    },
    {
      "entropy": 5.281605195999146,
      "epoch": 1.0975024015369836,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004887427359590897,
      "loss": 4.9057,
      "mean_token_accuracy": 0.21780946254730224,
      "num_tokens": 26182888.0,
      "step": 11425
    },
    {
      "entropy": 5.258964252471924,
      "epoch": 1.0979827089337175,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004887320250742482,
      "loss": 4.904,
      "mean_token_accuracy": 0.2275982156395912,
      "num_tokens": 26194979.0,
      "step": 11430
    },
    {
      "entropy": 5.212439107894897,
      "epoch": 1.0984630163304514,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004887213092271838,
      "loss": 4.9316,
      "mean_token_accuracy": 0.2295351058244705,
      "num_tokens": 26207309.0,
      "step": 11435
    },
    {
      "entropy": 5.269996070861817,
      "epoch": 1.0989433237271853,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004887105884181451,
      "loss": 4.9475,
      "mean_token_accuracy": 0.22447143346071244,
      "num_tokens": 26219231.0,
      "step": 11440
    },
    {
      "entropy": 5.346383047103882,
      "epoch": 1.0994236311239194,
      "grad_norm": 1.0,
      "learning_rate": 0.0004886998626473813,
      "loss": 4.9757,
      "mean_token_accuracy": 0.22025853097438813,
      "num_tokens": 26229355.0,
      "step": 11445
    },
    {
      "entropy": 5.3434593200683596,
      "epoch": 1.0999039385206533,
      "grad_norm": 1.0,
      "learning_rate": 0.0004886891319151411,
      "loss": 4.8814,
      "mean_token_accuracy": 0.22315512746572494,
      "num_tokens": 26239069.0,
      "step": 11450
    },
    {
      "entropy": 5.328772830963135,
      "epoch": 1.1003842459173871,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004886783962216738,
      "loss": 5.0178,
      "mean_token_accuracy": 0.21609151512384414,
      "num_tokens": 26250403.0,
      "step": 11455
    },
    {
      "entropy": 5.329624271392822,
      "epoch": 1.100864553314121,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004886676555672287,
      "loss": 4.8978,
      "mean_token_accuracy": 0.22373137176036834,
      "num_tokens": 26262926.0,
      "step": 11460
    },
    {
      "entropy": 5.360392236709595,
      "epoch": 1.101344860710855,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004886569099520551,
      "loss": 5.0397,
      "mean_token_accuracy": 0.21000013500452042,
      "num_tokens": 26274030.0,
      "step": 11465
    },
    {
      "entropy": 5.21960039138794,
      "epoch": 1.1018251681075888,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004886461593764024,
      "loss": 4.9431,
      "mean_token_accuracy": 0.2265513002872467,
      "num_tokens": 26284799.0,
      "step": 11470
    },
    {
      "entropy": 5.1629722118377686,
      "epoch": 1.1023054755043227,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004886354038405204,
      "loss": 4.8077,
      "mean_token_accuracy": 0.23654971569776534,
      "num_tokens": 26295154.0,
      "step": 11475
    },
    {
      "entropy": 5.332415056228638,
      "epoch": 1.1027857829010568,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004886246433446586,
      "loss": 4.9563,
      "mean_token_accuracy": 0.21839469820261,
      "num_tokens": 26306181.0,
      "step": 11480
    },
    {
      "entropy": 5.3780663967132565,
      "epoch": 1.1032660902977907,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004886138778890669,
      "loss": 5.0238,
      "mean_token_accuracy": 0.2137702763080597,
      "num_tokens": 26318674.0,
      "step": 11485
    },
    {
      "entropy": 5.31785020828247,
      "epoch": 1.1037463976945245,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004886031074739953,
      "loss": 4.9566,
      "mean_token_accuracy": 0.2117707684636116,
      "num_tokens": 26330257.0,
      "step": 11490
    },
    {
      "entropy": 5.294689035415649,
      "epoch": 1.1042267050912584,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004885923320996938,
      "loss": 4.8973,
      "mean_token_accuracy": 0.2266629695892334,
      "num_tokens": 26342035.0,
      "step": 11495
    },
    {
      "entropy": 5.292949914932251,
      "epoch": 1.1047070124879923,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004885815517664127,
      "loss": 4.958,
      "mean_token_accuracy": 0.22124435156583785,
      "num_tokens": 26352703.0,
      "step": 11500
    },
    {
      "entropy": 5.334151411056519,
      "epoch": 1.1051873198847262,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000488570766474402,
      "loss": 4.9622,
      "mean_token_accuracy": 0.22269196659326554,
      "num_tokens": 26362324.0,
      "step": 11505
    },
    {
      "entropy": 5.315319204330445,
      "epoch": 1.10566762728146,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004885599762239124,
      "loss": 4.9708,
      "mean_token_accuracy": 0.21237752437591553,
      "num_tokens": 26373540.0,
      "step": 11510
    },
    {
      "entropy": 5.235826730728149,
      "epoch": 1.106147934678194,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004885491810151943,
      "loss": 4.945,
      "mean_token_accuracy": 0.22126274853944777,
      "num_tokens": 26384259.0,
      "step": 11515
    },
    {
      "entropy": 5.222486782073974,
      "epoch": 1.106628242074928,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004885383808484982,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22843059450387954,
      "num_tokens": 26395193.0,
      "step": 11520
    },
    {
      "entropy": 5.248849582672119,
      "epoch": 1.107108549471662,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004885275757240751,
      "loss": 4.8733,
      "mean_token_accuracy": 0.2251675844192505,
      "num_tokens": 26408556.0,
      "step": 11525
    },
    {
      "entropy": 5.262370347976685,
      "epoch": 1.1075888568683958,
      "grad_norm": 1.0,
      "learning_rate": 0.0004885167656421757,
      "loss": 4.9394,
      "mean_token_accuracy": 0.22392557114362716,
      "num_tokens": 26420066.0,
      "step": 11530
    },
    {
      "entropy": 5.258542156219482,
      "epoch": 1.1080691642651297,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000488505950603051,
      "loss": 4.9004,
      "mean_token_accuracy": 0.22467269599437714,
      "num_tokens": 26432533.0,
      "step": 11535
    },
    {
      "entropy": 5.235738945007324,
      "epoch": 1.1085494716618636,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000488495130606952,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22229475378990174,
      "num_tokens": 26443878.0,
      "step": 11540
    },
    {
      "entropy": 5.279211139678955,
      "epoch": 1.1090297790585975,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004884843056541302,
      "loss": 4.9042,
      "mean_token_accuracy": 0.2213334932923317,
      "num_tokens": 26455111.0,
      "step": 11545
    },
    {
      "entropy": 5.261305141448974,
      "epoch": 1.1095100864553313,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004884734757448367,
      "loss": 4.9418,
      "mean_token_accuracy": 0.22275453060865402,
      "num_tokens": 26466577.0,
      "step": 11550
    },
    {
      "entropy": 5.255706834793091,
      "epoch": 1.1099903938520654,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000488462640879323,
      "loss": 4.8601,
      "mean_token_accuracy": 0.2271643027663231,
      "num_tokens": 26479637.0,
      "step": 11555
    },
    {
      "entropy": 5.233619260787964,
      "epoch": 1.1104707012487993,
      "grad_norm": 1.0,
      "learning_rate": 0.0004884518010578405,
      "loss": 4.8967,
      "mean_token_accuracy": 0.22283997684717177,
      "num_tokens": 26489923.0,
      "step": 11560
    },
    {
      "entropy": 5.3100982189178465,
      "epoch": 1.1109510086455332,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004884409562806411,
      "loss": 5.0052,
      "mean_token_accuracy": 0.22034398466348648,
      "num_tokens": 26501236.0,
      "step": 11565
    },
    {
      "entropy": 5.278788232803345,
      "epoch": 1.111431316042267,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004884301065479765,
      "loss": 4.862,
      "mean_token_accuracy": 0.2237830191850662,
      "num_tokens": 26511509.0,
      "step": 11570
    },
    {
      "entropy": 5.194205284118652,
      "epoch": 1.111911623439001,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004884192518600986,
      "loss": 4.8365,
      "mean_token_accuracy": 0.23374746441841127,
      "num_tokens": 26520931.0,
      "step": 11575
    },
    {
      "entropy": 5.179380798339844,
      "epoch": 1.1123919308357348,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004884083922172593,
      "loss": 4.8504,
      "mean_token_accuracy": 0.2359264761209488,
      "num_tokens": 26531333.0,
      "step": 11580
    },
    {
      "entropy": 5.210565614700317,
      "epoch": 1.1128722382324687,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004883975276197108,
      "loss": 4.9228,
      "mean_token_accuracy": 0.2243412986397743,
      "num_tokens": 26543696.0,
      "step": 11585
    },
    {
      "entropy": 5.24259386062622,
      "epoch": 1.1133525456292026,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004883866580677055,
      "loss": 4.8715,
      "mean_token_accuracy": 0.23089002817869186,
      "num_tokens": 26556292.0,
      "step": 11590
    },
    {
      "entropy": 5.2970452308654785,
      "epoch": 1.1138328530259365,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004883757835614956,
      "loss": 4.9021,
      "mean_token_accuracy": 0.21729104816913605,
      "num_tokens": 26568023.0,
      "step": 11595
    },
    {
      "entropy": 5.217583513259887,
      "epoch": 1.1143131604226706,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004883649041013335,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2274065524339676,
      "num_tokens": 26579469.0,
      "step": 11600
    },
    {
      "entropy": 5.3248392105102536,
      "epoch": 1.1147934678194045,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000488354019687472,
      "loss": 4.9281,
      "mean_token_accuracy": 0.22370994985103607,
      "num_tokens": 26591592.0,
      "step": 11605
    },
    {
      "entropy": 5.261428737640381,
      "epoch": 1.1152737752161384,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004883431303201636,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22833579182624816,
      "num_tokens": 26604075.0,
      "step": 11610
    },
    {
      "entropy": 5.188773155212402,
      "epoch": 1.1157540826128722,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004883322359996613,
      "loss": 4.8262,
      "mean_token_accuracy": 0.2327359840273857,
      "num_tokens": 26616448.0,
      "step": 11615
    },
    {
      "entropy": 5.280482482910156,
      "epoch": 1.1162343900096061,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004883213367262179,
      "loss": 4.9105,
      "mean_token_accuracy": 0.23350363671779634,
      "num_tokens": 26628749.0,
      "step": 11620
    },
    {
      "entropy": 5.232538509368896,
      "epoch": 1.11671469740634,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004883104325000866,
      "loss": 4.8761,
      "mean_token_accuracy": 0.22871393710374832,
      "num_tokens": 26640163.0,
      "step": 11625
    },
    {
      "entropy": 5.295828914642334,
      "epoch": 1.1171950048030739,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004882995233215203,
      "loss": 4.93,
      "mean_token_accuracy": 0.21729715764522553,
      "num_tokens": 26650729.0,
      "step": 11630
    },
    {
      "entropy": 5.261002254486084,
      "epoch": 1.117675312199808,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004882886091907726,
      "loss": 4.9463,
      "mean_token_accuracy": 0.22469919174909592,
      "num_tokens": 26661995.0,
      "step": 11635
    },
    {
      "entropy": 5.239017963409424,
      "epoch": 1.1181556195965419,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048827769010809666,
      "loss": 4.8679,
      "mean_token_accuracy": 0.2321700006723404,
      "num_tokens": 26673224.0,
      "step": 11640
    },
    {
      "entropy": 5.248235893249512,
      "epoch": 1.1186359269932757,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048826676607374606,
      "loss": 4.8952,
      "mean_token_accuracy": 0.22265169471502305,
      "num_tokens": 26686331.0,
      "step": 11645
    },
    {
      "entropy": 5.360962724685669,
      "epoch": 1.1191162343900096,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048825583708797434,
      "loss": 4.9566,
      "mean_token_accuracy": 0.22242544442415238,
      "num_tokens": 26696816.0,
      "step": 11650
    },
    {
      "entropy": 5.2435211658477785,
      "epoch": 1.1195965417867435,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004882449031510354,
      "loss": 4.9009,
      "mean_token_accuracy": 0.22770000994205475,
      "num_tokens": 26708126.0,
      "step": 11655
    },
    {
      "entropy": 5.317254734039307,
      "epoch": 1.1200768491834774,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000488233964263183,
      "loss": 5.0156,
      "mean_token_accuracy": 0.21930047273635864,
      "num_tokens": 26718951.0,
      "step": 11660
    },
    {
      "entropy": 5.309821128845215,
      "epoch": 1.1205571565802113,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048822302042467115,
      "loss": 4.9769,
      "mean_token_accuracy": 0.22061043381690978,
      "num_tokens": 26730550.0,
      "step": 11665
    },
    {
      "entropy": 5.216285467147827,
      "epoch": 1.1210374639769451,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004882120716357539,
      "loss": 4.8174,
      "mean_token_accuracy": 0.23178662806749345,
      "num_tokens": 26741485.0,
      "step": 11670
    },
    {
      "entropy": 5.167406034469605,
      "epoch": 1.1215177713736793,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004882011178966854,
      "loss": 4.8108,
      "mean_token_accuracy": 0.22928302437067033,
      "num_tokens": 26753947.0,
      "step": 11675
    },
    {
      "entropy": 5.24999852180481,
      "epoch": 1.1219980787704131,
      "grad_norm": 1.0,
      "learning_rate": 0.0004881901592077201,
      "loss": 4.8742,
      "mean_token_accuracy": 0.23690653443336487,
      "num_tokens": 26764921.0,
      "step": 11680
    },
    {
      "entropy": 5.2873766899108885,
      "epoch": 1.122478386167147,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004881791955691123,
      "loss": 4.8523,
      "mean_token_accuracy": 0.22306306213140487,
      "num_tokens": 26777384.0,
      "step": 11685
    },
    {
      "entropy": 5.272368001937866,
      "epoch": 1.122958693563881,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048816822698111655,
      "loss": 4.9601,
      "mean_token_accuracy": 0.21977699995040895,
      "num_tokens": 26789916.0,
      "step": 11690
    },
    {
      "entropy": 5.241760158538819,
      "epoch": 1.1234390009606148,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048815725344398766,
      "loss": 4.8673,
      "mean_token_accuracy": 0.22854420244693757,
      "num_tokens": 26801167.0,
      "step": 11695
    },
    {
      "entropy": 5.332319498062134,
      "epoch": 1.1239193083573487,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048814627495798017,
      "loss": 4.9657,
      "mean_token_accuracy": 0.22256825715303422,
      "num_tokens": 26813235.0,
      "step": 11700
    },
    {
      "entropy": 5.31646237373352,
      "epoch": 1.1243996157540825,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004881352915233492,
      "loss": 4.9644,
      "mean_token_accuracy": 0.22639967501163483,
      "num_tokens": 26824758.0,
      "step": 11705
    },
    {
      "entropy": 5.286846160888672,
      "epoch": 1.1248799231508164,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048812430314034956,
      "loss": 4.9409,
      "mean_token_accuracy": 0.21787350177764891,
      "num_tokens": 26836877.0,
      "step": 11710
    },
    {
      "entropy": 5.173878479003906,
      "epoch": 1.1253602305475505,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004881133098092365,
      "loss": 4.7838,
      "mean_token_accuracy": 0.23808563351631165,
      "num_tokens": 26848394.0,
      "step": 11715
    },
    {
      "entropy": 5.243302726745606,
      "epoch": 1.1258405379442844,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004881023115302652,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22353948652744293,
      "num_tokens": 26859064.0,
      "step": 11720
    },
    {
      "entropy": 5.342434310913086,
      "epoch": 1.1263208453410183,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000488091308303691,
      "loss": 5.0307,
      "mean_token_accuracy": 0.21131904274225236,
      "num_tokens": 26870401.0,
      "step": 11725
    },
    {
      "entropy": 5.302391052246094,
      "epoch": 1.1268011527377522,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004880803001297694,
      "loss": 4.9226,
      "mean_token_accuracy": 0.22648945450782776,
      "num_tokens": 26881767.0,
      "step": 11730
    },
    {
      "entropy": 5.230623817443847,
      "epoch": 1.127281460134486,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004880692870087559,
      "loss": 4.8339,
      "mean_token_accuracy": 0.22091956436634064,
      "num_tokens": 26892854.0,
      "step": 11735
    },
    {
      "entropy": 5.373048162460327,
      "epoch": 1.12776176753122,
      "grad_norm": 1.0,
      "learning_rate": 0.00048805826894090626,
      "loss": 4.9983,
      "mean_token_accuracy": 0.2149614080786705,
      "num_tokens": 26905079.0,
      "step": 11740
    },
    {
      "entropy": 5.365033292770386,
      "epoch": 1.1282420749279538,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048804724592647626,
      "loss": 5.0227,
      "mean_token_accuracy": 0.21428083777427673,
      "num_tokens": 26915640.0,
      "step": 11745
    },
    {
      "entropy": 5.280916404724121,
      "epoch": 1.1287223823246877,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004880362179657218,
      "loss": 4.9048,
      "mean_token_accuracy": 0.22199257612228393,
      "num_tokens": 26927039.0,
      "step": 11750
    },
    {
      "entropy": 5.23762321472168,
      "epoch": 1.1292026897214218,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048802518505889904,
      "loss": 4.8817,
      "mean_token_accuracy": 0.22747842967510223,
      "num_tokens": 26939368.0,
      "step": 11755
    },
    {
      "entropy": 5.241123676300049,
      "epoch": 1.1296829971181557,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048801414720626404,
      "loss": 5.0453,
      "mean_token_accuracy": 0.2083485558629036,
      "num_tokens": 26952051.0,
      "step": 11760
    },
    {
      "entropy": 5.248374652862549,
      "epoch": 1.1301633045148896,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048800310440807294,
      "loss": 4.8854,
      "mean_token_accuracy": 0.22071474939584732,
      "num_tokens": 26963680.0,
      "step": 11765
    },
    {
      "entropy": 5.172375965118408,
      "epoch": 1.1306436119116234,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004879920566645823,
      "loss": 4.7888,
      "mean_token_accuracy": 0.23066822588443756,
      "num_tokens": 26973899.0,
      "step": 11770
    },
    {
      "entropy": 5.22331748008728,
      "epoch": 1.1311239193083573,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004879810039760486,
      "loss": 4.9262,
      "mean_token_accuracy": 0.21739015877246856,
      "num_tokens": 26985692.0,
      "step": 11775
    },
    {
      "entropy": 5.29095311164856,
      "epoch": 1.1316042267050912,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004879699463427284,
      "loss": 4.8715,
      "mean_token_accuracy": 0.22801961600780488,
      "num_tokens": 26996354.0,
      "step": 11780
    },
    {
      "entropy": 5.3187196254730225,
      "epoch": 1.132084534101825,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004879588837648785,
      "loss": 4.903,
      "mean_token_accuracy": 0.2252940446138382,
      "num_tokens": 27007925.0,
      "step": 11785
    },
    {
      "entropy": 5.229143381118774,
      "epoch": 1.1325648414985592,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048794781624275554,
      "loss": 4.8719,
      "mean_token_accuracy": 0.2319113165140152,
      "num_tokens": 27019222.0,
      "step": 11790
    },
    {
      "entropy": 5.186465501785278,
      "epoch": 1.133045148895293,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048793674377661664,
      "loss": 4.8715,
      "mean_token_accuracy": 0.23309751152992247,
      "num_tokens": 27029973.0,
      "step": 11795
    },
    {
      "entropy": 5.31542739868164,
      "epoch": 1.133525456292027,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048792566636671886,
      "loss": 5.0043,
      "mean_token_accuracy": 0.2228681653738022,
      "num_tokens": 27042002.0,
      "step": 11800
    },
    {
      "entropy": 5.287324142456055,
      "epoch": 1.1340057636887608,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004879145840133194,
      "loss": 4.9072,
      "mean_token_accuracy": 0.23078465461730957,
      "num_tokens": 27052770.0,
      "step": 11805
    },
    {
      "entropy": 5.295166921615601,
      "epoch": 1.1344860710854947,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004879034967166755,
      "loss": 4.9692,
      "mean_token_accuracy": 0.22118282169103623,
      "num_tokens": 27063375.0,
      "step": 11810
    },
    {
      "entropy": 5.197003650665283,
      "epoch": 1.1349663784822286,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004878924044770446,
      "loss": 4.9009,
      "mean_token_accuracy": 0.22608220279216767,
      "num_tokens": 27075908.0,
      "step": 11815
    },
    {
      "entropy": 5.265542221069336,
      "epoch": 1.1354466858789625,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004878813072946843,
      "loss": 4.9377,
      "mean_token_accuracy": 0.22460416406393052,
      "num_tokens": 27087590.0,
      "step": 11820
    },
    {
      "entropy": 5.269877195358276,
      "epoch": 1.1359269932756964,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048787020516985203,
      "loss": 4.8576,
      "mean_token_accuracy": 0.23368470519781112,
      "num_tokens": 27098076.0,
      "step": 11825
    },
    {
      "entropy": 5.228240251541138,
      "epoch": 1.1364073006724302,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048785909810280576,
      "loss": 4.8673,
      "mean_token_accuracy": 0.23402309268712998,
      "num_tokens": 27109851.0,
      "step": 11830
    },
    {
      "entropy": 5.248457574844361,
      "epoch": 1.1368876080691643,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004878479860938033,
      "loss": 4.8227,
      "mean_token_accuracy": 0.23155975639820098,
      "num_tokens": 27121288.0,
      "step": 11835
    },
    {
      "entropy": 5.263022136688233,
      "epoch": 1.1373679154658982,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048783686914310266,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22765712738037108,
      "num_tokens": 27133877.0,
      "step": 11840
    },
    {
      "entropy": 5.113182592391968,
      "epoch": 1.137848222862632,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004878257472509619,
      "loss": 4.7205,
      "mean_token_accuracy": 0.23631683439016343,
      "num_tokens": 27145616.0,
      "step": 11845
    },
    {
      "entropy": 5.19237904548645,
      "epoch": 1.138328530259366,
      "grad_norm": 1.0,
      "learning_rate": 0.0004878146204176392,
      "loss": 4.8815,
      "mean_token_accuracy": 0.22343646436929704,
      "num_tokens": 27157182.0,
      "step": 11850
    },
    {
      "entropy": 5.28602442741394,
      "epoch": 1.1388088376560999,
      "grad_norm": 1.015625,
      "learning_rate": 0.000487803488643393,
      "loss": 4.839,
      "mean_token_accuracy": 0.2278805673122406,
      "num_tokens": 27167609.0,
      "step": 11855
    },
    {
      "entropy": 5.320325565338135,
      "epoch": 1.1392891450528337,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048779235192848166,
      "loss": 4.9875,
      "mean_token_accuracy": 0.21549776196479797,
      "num_tokens": 27180043.0,
      "step": 11860
    },
    {
      "entropy": 5.209945821762085,
      "epoch": 1.1397694524495678,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004877812102731638,
      "loss": 4.7885,
      "mean_token_accuracy": 0.2342821851372719,
      "num_tokens": 27191870.0,
      "step": 11865
    },
    {
      "entropy": 5.231264543533325,
      "epoch": 1.1402497598463017,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048777006367769804,
      "loss": 4.9274,
      "mean_token_accuracy": 0.2248838245868683,
      "num_tokens": 27203291.0,
      "step": 11870
    },
    {
      "entropy": 5.248008155822754,
      "epoch": 1.1407300672430356,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004877589121423432,
      "loss": 4.9359,
      "mean_token_accuracy": 0.22610438615083694,
      "num_tokens": 27214607.0,
      "step": 11875
    },
    {
      "entropy": 5.247781038284302,
      "epoch": 1.1412103746397695,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004877477556673582,
      "loss": 4.8043,
      "mean_token_accuracy": 0.23223898261785508,
      "num_tokens": 27224658.0,
      "step": 11880
    },
    {
      "entropy": 5.188546466827392,
      "epoch": 1.1416906820365034,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000487736594253002,
      "loss": 4.9207,
      "mean_token_accuracy": 0.21762831658124923,
      "num_tokens": 27235306.0,
      "step": 11885
    },
    {
      "entropy": 5.2809553146362305,
      "epoch": 1.1421709894332372,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048772542789953384,
      "loss": 4.896,
      "mean_token_accuracy": 0.2212103322148323,
      "num_tokens": 27246679.0,
      "step": 11890
    },
    {
      "entropy": 5.352729797363281,
      "epoch": 1.1426512968299711,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048771425660721284,
      "loss": 4.9452,
      "mean_token_accuracy": 0.21574013829231262,
      "num_tokens": 27257799.0,
      "step": 11895
    },
    {
      "entropy": 5.22221736907959,
      "epoch": 1.143131604226705,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048770308037629853,
      "loss": 4.9424,
      "mean_token_accuracy": 0.22030197083950043,
      "num_tokens": 27268651.0,
      "step": 11900
    },
    {
      "entropy": 5.204977560043335,
      "epoch": 1.143611911623439,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004876918992070502,
      "loss": 4.9007,
      "mean_token_accuracy": 0.22769503444433212,
      "num_tokens": 27280655.0,
      "step": 11905
    },
    {
      "entropy": 5.3163105010986325,
      "epoch": 1.144092219020173,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004876807130997276,
      "loss": 4.9744,
      "mean_token_accuracy": 0.2218126729130745,
      "num_tokens": 27292664.0,
      "step": 11910
    },
    {
      "entropy": 5.2656238079071045,
      "epoch": 1.1445725264169069,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004876695220545903,
      "loss": 4.9476,
      "mean_token_accuracy": 0.22341082394123077,
      "num_tokens": 27306143.0,
      "step": 11915
    },
    {
      "entropy": 5.255891180038452,
      "epoch": 1.1450528338136408,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048765832607189824,
      "loss": 4.8455,
      "mean_token_accuracy": 0.22855520248413086,
      "num_tokens": 27316926.0,
      "step": 11920
    },
    {
      "entropy": 5.333289289474488,
      "epoch": 1.1455331412103746,
      "grad_norm": 1.0,
      "learning_rate": 0.00048764712515191136,
      "loss": 4.9626,
      "mean_token_accuracy": 0.22719553709030152,
      "num_tokens": 27327472.0,
      "step": 11925
    },
    {
      "entropy": 5.311666631698609,
      "epoch": 1.1460134486071085,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048763591929488966,
      "loss": 4.9464,
      "mean_token_accuracy": 0.22794317007064818,
      "num_tokens": 27338388.0,
      "step": 11930
    },
    {
      "entropy": 5.236239624023438,
      "epoch": 1.1464937560038424,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004876247085010933,
      "loss": 4.8448,
      "mean_token_accuracy": 0.2247237354516983,
      "num_tokens": 27349879.0,
      "step": 11935
    },
    {
      "entropy": 5.207219505310059,
      "epoch": 1.1469740634005763,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048761349277078253,
      "loss": 4.8669,
      "mean_token_accuracy": 0.23082393556833267,
      "num_tokens": 27361226.0,
      "step": 11940
    },
    {
      "entropy": 5.249807929992675,
      "epoch": 1.1474543707973104,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048760227210421775,
      "loss": 4.8419,
      "mean_token_accuracy": 0.22972375750541688,
      "num_tokens": 27373512.0,
      "step": 11945
    },
    {
      "entropy": 5.202455377578735,
      "epoch": 1.1479346781940443,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004875910465016596,
      "loss": 4.9001,
      "mean_token_accuracy": 0.22783205211162566,
      "num_tokens": 27384950.0,
      "step": 11950
    },
    {
      "entropy": 5.370511722564697,
      "epoch": 1.1484149855907781,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004875798159633686,
      "loss": 4.9569,
      "mean_token_accuracy": 0.22006280571222306,
      "num_tokens": 27395877.0,
      "step": 11955
    },
    {
      "entropy": 5.287519121170044,
      "epoch": 1.148895292987512,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004875685804896055,
      "loss": 4.9874,
      "mean_token_accuracy": 0.22074266523122787,
      "num_tokens": 27407783.0,
      "step": 11960
    },
    {
      "entropy": 5.218634033203125,
      "epoch": 1.149375600384246,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004875573400806312,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22635357081890106,
      "num_tokens": 27420598.0,
      "step": 11965
    },
    {
      "entropy": 5.277797985076904,
      "epoch": 1.1498559077809798,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048754609473670654,
      "loss": 4.9551,
      "mean_token_accuracy": 0.2221503108739853,
      "num_tokens": 27431961.0,
      "step": 11970
    },
    {
      "entropy": 5.240969467163086,
      "epoch": 1.1503362151777137,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004875348444580927,
      "loss": 4.8152,
      "mean_token_accuracy": 0.2302515208721161,
      "num_tokens": 27444010.0,
      "step": 11975
    },
    {
      "entropy": 5.302516269683838,
      "epoch": 1.1508165225744476,
      "grad_norm": 0.921875,
      "learning_rate": 0.000487523589245051,
      "loss": 4.8866,
      "mean_token_accuracy": 0.22645665407180787,
      "num_tokens": 27455576.0,
      "step": 11980
    },
    {
      "entropy": 5.267692232131958,
      "epoch": 1.1512968299711814,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004875123290978425,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2249714821577072,
      "num_tokens": 27468957.0,
      "step": 11985
    },
    {
      "entropy": 5.243158531188965,
      "epoch": 1.1517771373679155,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048750106401672876,
      "loss": 4.9001,
      "mean_token_accuracy": 0.23189364224672318,
      "num_tokens": 27479992.0,
      "step": 11990
    },
    {
      "entropy": 5.305775213241577,
      "epoch": 1.1522574447646494,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048748979400197134,
      "loss": 4.9592,
      "mean_token_accuracy": 0.21976903229951858,
      "num_tokens": 27490878.0,
      "step": 11995
    },
    {
      "entropy": 5.204423952102661,
      "epoch": 1.1527377521613833,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048747851905383183,
      "loss": 4.8139,
      "mean_token_accuracy": 0.2335945963859558,
      "num_tokens": 27502009.0,
      "step": 12000
    },
    {
      "epoch": 1.1527377521613833,
      "eval_entropy": 5.021074163222069,
      "eval_loss": 4.997524738311768,
      "eval_mean_token_accuracy": 0.2287947643795271,
      "eval_num_tokens": 27502009.0,
      "eval_runtime": 20.5513,
      "eval_samples_per_second": 1596.734,
      "eval_steps_per_second": 199.598,
      "step": 12000
    },
    {
      "entropy": 5.255040693283081,
      "epoch": 1.1532180595581172,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004874672391725721,
      "loss": 4.9903,
      "mean_token_accuracy": 0.21744824200868607,
      "num_tokens": 27513376.0,
      "step": 12005
    },
    {
      "entropy": 5.192897796630859,
      "epoch": 1.153698366954851,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004874559543584539,
      "loss": 4.851,
      "mean_token_accuracy": 0.22504581212997438,
      "num_tokens": 27525166.0,
      "step": 12010
    },
    {
      "entropy": 5.203159284591675,
      "epoch": 1.154178674351585,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004874446646117394,
      "loss": 4.7842,
      "mean_token_accuracy": 0.2401562139391899,
      "num_tokens": 27535994.0,
      "step": 12015
    },
    {
      "entropy": 5.231074666976928,
      "epoch": 1.154658981748319,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004874333699326906,
      "loss": 4.8724,
      "mean_token_accuracy": 0.22671962529420853,
      "num_tokens": 27546883.0,
      "step": 12020
    },
    {
      "entropy": 5.319095754623413,
      "epoch": 1.155139289145053,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004874220703215697,
      "loss": 4.9834,
      "mean_token_accuracy": 0.21863362193107605,
      "num_tokens": 27558840.0,
      "step": 12025
    },
    {
      "entropy": 5.263051128387451,
      "epoch": 1.1556195965417868,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004874107657786391,
      "loss": 4.9077,
      "mean_token_accuracy": 0.22483978420495987,
      "num_tokens": 27569776.0,
      "step": 12030
    },
    {
      "entropy": 5.285828399658203,
      "epoch": 1.1560999039385207,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048739945630416124,
      "loss": 4.8569,
      "mean_token_accuracy": 0.2249127522110939,
      "num_tokens": 27580097.0,
      "step": 12035
    },
    {
      "entropy": 5.301922416687011,
      "epoch": 1.1565802113352546,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004873881418983987,
      "loss": 4.9959,
      "mean_token_accuracy": 0.22298937290906906,
      "num_tokens": 27592035.0,
      "step": 12040
    },
    {
      "entropy": 5.308622789382935,
      "epoch": 1.1570605187319885,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004873768225616141,
      "loss": 4.9384,
      "mean_token_accuracy": 0.22553286403417588,
      "num_tokens": 27604409.0,
      "step": 12045
    },
    {
      "entropy": 5.281864738464355,
      "epoch": 1.1575408261287223,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048736549829407047,
      "loss": 4.8712,
      "mean_token_accuracy": 0.2261851966381073,
      "num_tokens": 27615019.0,
      "step": 12050
    },
    {
      "entropy": 5.295188570022583,
      "epoch": 1.1580211335254562,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004873541690960305,
      "loss": 5.0102,
      "mean_token_accuracy": 0.21841889768838882,
      "num_tokens": 27627924.0,
      "step": 12055
    },
    {
      "entropy": 5.232680368423462,
      "epoch": 1.15850144092219,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004873428349677573,
      "loss": 4.817,
      "mean_token_accuracy": 0.22325704395771026,
      "num_tokens": 27639188.0,
      "step": 12060
    },
    {
      "entropy": 5.356967973709106,
      "epoch": 1.1589817483189242,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000487331495909514,
      "loss": 5.0325,
      "mean_token_accuracy": 0.2123776152729988,
      "num_tokens": 27652621.0,
      "step": 12065
    },
    {
      "entropy": 5.288081359863281,
      "epoch": 1.159462055715658,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048732015192156383,
      "loss": 4.9518,
      "mean_token_accuracy": 0.21672327518463136,
      "num_tokens": 27665726.0,
      "step": 12070
    },
    {
      "entropy": 5.265194463729858,
      "epoch": 1.159942363112392,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048730880300417015,
      "loss": 4.9192,
      "mean_token_accuracy": 0.21923461258411409,
      "num_tokens": 27676984.0,
      "step": 12075
    },
    {
      "entropy": 5.310585832595825,
      "epoch": 1.1604226705091258,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048729744915759657,
      "loss": 4.9708,
      "mean_token_accuracy": 0.22491427809000014,
      "num_tokens": 27688812.0,
      "step": 12080
    },
    {
      "entropy": 5.186745309829712,
      "epoch": 1.1609029779058597,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048728609038210655,
      "loss": 4.7839,
      "mean_token_accuracy": 0.2299763321876526,
      "num_tokens": 27699152.0,
      "step": 12085
    },
    {
      "entropy": 5.245731592178345,
      "epoch": 1.1613832853025936,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048727472667796395,
      "loss": 4.9508,
      "mean_token_accuracy": 0.2212846577167511,
      "num_tokens": 27710347.0,
      "step": 12090
    },
    {
      "entropy": 5.212558555603027,
      "epoch": 1.1618635926993275,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004872633580454325,
      "loss": 4.8605,
      "mean_token_accuracy": 0.23117222785949706,
      "num_tokens": 27721112.0,
      "step": 12095
    },
    {
      "entropy": 5.306494808197021,
      "epoch": 1.1623439000960616,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048725198448477616,
      "loss": 4.95,
      "mean_token_accuracy": 0.2186781033873558,
      "num_tokens": 27731766.0,
      "step": 12100
    },
    {
      "entropy": 5.231584310531616,
      "epoch": 1.1628242074927955,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048724060599625893,
      "loss": 4.8943,
      "mean_token_accuracy": 0.22573711425065995,
      "num_tokens": 27743718.0,
      "step": 12105
    },
    {
      "entropy": 5.257784128189087,
      "epoch": 1.1633045148895294,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048722922258014506,
      "loss": 4.9414,
      "mean_token_accuracy": 0.22455735504627228,
      "num_tokens": 27754999.0,
      "step": 12110
    },
    {
      "entropy": 5.330732011795044,
      "epoch": 1.1637848222862632,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004872178342366989,
      "loss": 5.0293,
      "mean_token_accuracy": 0.2158423110842705,
      "num_tokens": 27767684.0,
      "step": 12115
    },
    {
      "entropy": 5.250082635879517,
      "epoch": 1.1642651296829971,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048720644096618475,
      "loss": 4.9008,
      "mean_token_accuracy": 0.22699017375707625,
      "num_tokens": 27779103.0,
      "step": 12120
    },
    {
      "entropy": 5.262233686447144,
      "epoch": 1.164745437079731,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004871950427688672,
      "loss": 4.9297,
      "mean_token_accuracy": 0.2294217824935913,
      "num_tokens": 27791223.0,
      "step": 12125
    },
    {
      "entropy": 5.275623178482055,
      "epoch": 1.1652257444764649,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048718363964501087,
      "loss": 4.8743,
      "mean_token_accuracy": 0.2276843160390854,
      "num_tokens": 27802660.0,
      "step": 12130
    },
    {
      "entropy": 5.2676025390625,
      "epoch": 1.1657060518731988,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004871722315948805,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21871267557144164,
      "num_tokens": 27814370.0,
      "step": 12135
    },
    {
      "entropy": 5.257337522506714,
      "epoch": 1.1661863592699326,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004871608186187408,
      "loss": 4.8468,
      "mean_token_accuracy": 0.23260480612516404,
      "num_tokens": 27825035.0,
      "step": 12140
    },
    {
      "entropy": 5.166463041305542,
      "epoch": 1.1666666666666667,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048714940071685703,
      "loss": 4.7855,
      "mean_token_accuracy": 0.23018673211336135,
      "num_tokens": 27837276.0,
      "step": 12145
    },
    {
      "entropy": 5.344819736480713,
      "epoch": 1.1671469740634006,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048713797788949405,
      "loss": 4.9656,
      "mean_token_accuracy": 0.21884248703718184,
      "num_tokens": 27848618.0,
      "step": 12150
    },
    {
      "entropy": 5.232809400558471,
      "epoch": 1.1676272814601345,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048712655013691714,
      "loss": 4.9256,
      "mean_token_accuracy": 0.21615913957357408,
      "num_tokens": 27861556.0,
      "step": 12155
    },
    {
      "entropy": 5.283457946777344,
      "epoch": 1.1681075888568684,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048711511745939165,
      "loss": 4.9091,
      "mean_token_accuracy": 0.2208642080426216,
      "num_tokens": 27873211.0,
      "step": 12160
    },
    {
      "entropy": 5.271178817749023,
      "epoch": 1.1685878962536023,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000487103679857183,
      "loss": 4.9118,
      "mean_token_accuracy": 0.22803781032562256,
      "num_tokens": 27883365.0,
      "step": 12165
    },
    {
      "entropy": 5.1629351615905765,
      "epoch": 1.1690682036503361,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004870922373305567,
      "loss": 4.8149,
      "mean_token_accuracy": 0.2291519120335579,
      "num_tokens": 27894669.0,
      "step": 12170
    },
    {
      "entropy": 5.232136058807373,
      "epoch": 1.1695485110470702,
      "grad_norm": 0.875,
      "learning_rate": 0.00048708078987977837,
      "loss": 4.9155,
      "mean_token_accuracy": 0.2255495697259903,
      "num_tokens": 27906480.0,
      "step": 12175
    },
    {
      "entropy": 5.247304105758667,
      "epoch": 1.1700288184438041,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048706933750511394,
      "loss": 4.8675,
      "mean_token_accuracy": 0.2324027270078659,
      "num_tokens": 27918383.0,
      "step": 12180
    },
    {
      "entropy": 5.329435873031616,
      "epoch": 1.170509125840538,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004870578802068292,
      "loss": 4.9239,
      "mean_token_accuracy": 0.22166726738214493,
      "num_tokens": 27928944.0,
      "step": 12185
    },
    {
      "entropy": 5.203683567047119,
      "epoch": 1.170989433237272,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048704641798519006,
      "loss": 4.8429,
      "mean_token_accuracy": 0.22740177661180497,
      "num_tokens": 27941105.0,
      "step": 12190
    },
    {
      "entropy": 5.251979970932007,
      "epoch": 1.1714697406340058,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048703495084046286,
      "loss": 4.88,
      "mean_token_accuracy": 0.230315200984478,
      "num_tokens": 27952925.0,
      "step": 12195
    },
    {
      "entropy": 5.186851549148559,
      "epoch": 1.1719500480307397,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004870234787729137,
      "loss": 4.8747,
      "mean_token_accuracy": 0.23460304141044616,
      "num_tokens": 27965504.0,
      "step": 12200
    },
    {
      "entropy": 5.2175195693969725,
      "epoch": 1.1724303554274735,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004870120017828089,
      "loss": 4.841,
      "mean_token_accuracy": 0.23686108738183975,
      "num_tokens": 27975985.0,
      "step": 12205
    },
    {
      "entropy": 5.357242250442505,
      "epoch": 1.1729106628242074,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000487000519870415,
      "loss": 4.961,
      "mean_token_accuracy": 0.21317480206489564,
      "num_tokens": 27987784.0,
      "step": 12210
    },
    {
      "entropy": 5.259235191345215,
      "epoch": 1.1733909702209413,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004869890330359986,
      "loss": 4.8759,
      "mean_token_accuracy": 0.22611529231071473,
      "num_tokens": 27999623.0,
      "step": 12215
    },
    {
      "entropy": 5.248368501663208,
      "epoch": 1.1738712776176754,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004869775412798262,
      "loss": 4.8942,
      "mean_token_accuracy": 0.22390271574258805,
      "num_tokens": 28011561.0,
      "step": 12220
    },
    {
      "entropy": 5.28819842338562,
      "epoch": 1.1743515850144093,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048696604460216476,
      "loss": 4.9251,
      "mean_token_accuracy": 0.23209764212369918,
      "num_tokens": 28022108.0,
      "step": 12225
    },
    {
      "entropy": 5.2934918880462645,
      "epoch": 1.1748318924111432,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048695454300328123,
      "loss": 4.9243,
      "mean_token_accuracy": 0.22135676741600036,
      "num_tokens": 28033497.0,
      "step": 12230
    },
    {
      "entropy": 5.313299369812012,
      "epoch": 1.175312199807877,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048694303648344256,
      "loss": 4.9368,
      "mean_token_accuracy": 0.22623614221811295,
      "num_tokens": 28044790.0,
      "step": 12235
    },
    {
      "entropy": 5.213284015655518,
      "epoch": 1.175792507204611,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048693152504291595,
      "loss": 4.8779,
      "mean_token_accuracy": 0.22745126336812974,
      "num_tokens": 28056000.0,
      "step": 12240
    },
    {
      "entropy": 5.2219573020935055,
      "epoch": 1.1762728146013448,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004869200086819686,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22550372481346131,
      "num_tokens": 28068264.0,
      "step": 12245
    },
    {
      "entropy": 5.295480680465698,
      "epoch": 1.1767531219980787,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048690848740086796,
      "loss": 4.8153,
      "mean_token_accuracy": 0.23126688003540039,
      "num_tokens": 28080220.0,
      "step": 12250
    },
    {
      "entropy": 5.303543519973755,
      "epoch": 1.1772334293948128,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004868969611998814,
      "loss": 4.8614,
      "mean_token_accuracy": 0.2277045503258705,
      "num_tokens": 28091531.0,
      "step": 12255
    },
    {
      "entropy": 5.218742990493775,
      "epoch": 1.1777137367915467,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004868854300792767,
      "loss": 4.8658,
      "mean_token_accuracy": 0.2224758580327034,
      "num_tokens": 28102982.0,
      "step": 12260
    },
    {
      "entropy": 5.2459714889526365,
      "epoch": 1.1781940441882806,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048687389403932144,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2236674413084984,
      "num_tokens": 28114131.0,
      "step": 12265
    },
    {
      "entropy": 5.256588506698608,
      "epoch": 1.1786743515850144,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004868623530802835,
      "loss": 4.8585,
      "mean_token_accuracy": 0.22192418873310088,
      "num_tokens": 28125637.0,
      "step": 12270
    },
    {
      "entropy": 5.344217300415039,
      "epoch": 1.1791546589817483,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048685080720243086,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2285078302025795,
      "num_tokens": 28137284.0,
      "step": 12275
    },
    {
      "entropy": 5.29886212348938,
      "epoch": 1.1796349663784822,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004868392564060315,
      "loss": 4.9294,
      "mean_token_accuracy": 0.22919966876506806,
      "num_tokens": 28148909.0,
      "step": 12280
    },
    {
      "entropy": 5.280670976638794,
      "epoch": 1.180115273775216,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004868277006913537,
      "loss": 4.9231,
      "mean_token_accuracy": 0.2273537814617157,
      "num_tokens": 28160392.0,
      "step": 12285
    },
    {
      "entropy": 5.2928205013275145,
      "epoch": 1.18059558117195,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004868161400586656,
      "loss": 4.978,
      "mean_token_accuracy": 0.2159964546561241,
      "num_tokens": 28172957.0,
      "step": 12290
    },
    {
      "entropy": 5.36279149055481,
      "epoch": 1.1810758885686838,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004868045745082357,
      "loss": 4.9786,
      "mean_token_accuracy": 0.21561757028102874,
      "num_tokens": 28183239.0,
      "step": 12295
    },
    {
      "entropy": 5.247126340866089,
      "epoch": 1.181556195965418,
      "grad_norm": 1.0,
      "learning_rate": 0.0004867930040403326,
      "loss": 4.8774,
      "mean_token_accuracy": 0.22056290805339812,
      "num_tokens": 28195382.0,
      "step": 12300
    },
    {
      "entropy": 5.251654815673828,
      "epoch": 1.1820365033621518,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048678142865522475,
      "loss": 4.942,
      "mean_token_accuracy": 0.21987020522356032,
      "num_tokens": 28206645.0,
      "step": 12305
    },
    {
      "entropy": 5.309177446365356,
      "epoch": 1.1825168107588857,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000486769848353181,
      "loss": 4.9352,
      "mean_token_accuracy": 0.22527330964803696,
      "num_tokens": 28218420.0,
      "step": 12310
    },
    {
      "entropy": 5.31207127571106,
      "epoch": 1.1829971181556196,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048675826313447027,
      "loss": 4.9533,
      "mean_token_accuracy": 0.22487905621528625,
      "num_tokens": 28229458.0,
      "step": 12315
    },
    {
      "entropy": 5.286565828323364,
      "epoch": 1.1834774255523535,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048674667299936135,
      "loss": 4.944,
      "mean_token_accuracy": 0.21893807053565978,
      "num_tokens": 28242240.0,
      "step": 12320
    },
    {
      "entropy": 5.418816661834716,
      "epoch": 1.1839577329490873,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048673507794812356,
      "loss": 5.0236,
      "mean_token_accuracy": 0.21421619951725007,
      "num_tokens": 28254597.0,
      "step": 12325
    },
    {
      "entropy": 5.262916040420532,
      "epoch": 1.1844380403458212,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004867234779810259,
      "loss": 4.9689,
      "mean_token_accuracy": 0.221517650783062,
      "num_tokens": 28266674.0,
      "step": 12330
    },
    {
      "entropy": 5.225225305557251,
      "epoch": 1.1849183477425553,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004867118730983378,
      "loss": 4.9056,
      "mean_token_accuracy": 0.23125376254320146,
      "num_tokens": 28278671.0,
      "step": 12335
    },
    {
      "entropy": 5.296021413803101,
      "epoch": 1.1853986551392892,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004867002633003286,
      "loss": 4.8466,
      "mean_token_accuracy": 0.22779525369405745,
      "num_tokens": 28291085.0,
      "step": 12340
    },
    {
      "entropy": 5.27785849571228,
      "epoch": 1.185878962536023,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000486688648587268,
      "loss": 4.9077,
      "mean_token_accuracy": 0.2296278804540634,
      "num_tokens": 28303457.0,
      "step": 12345
    },
    {
      "entropy": 5.285258388519287,
      "epoch": 1.186359269932757,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004866770289594256,
      "loss": 4.9377,
      "mean_token_accuracy": 0.2224503979086876,
      "num_tokens": 28314599.0,
      "step": 12350
    },
    {
      "entropy": 5.37945818901062,
      "epoch": 1.1868395773294909,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048666540441707107,
      "loss": 5.0028,
      "mean_token_accuracy": 0.22393649965524673,
      "num_tokens": 28326266.0,
      "step": 12355
    },
    {
      "entropy": 5.256625843048096,
      "epoch": 1.1873198847262247,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004866537749604744,
      "loss": 4.8788,
      "mean_token_accuracy": 0.22460722774267197,
      "num_tokens": 28337238.0,
      "step": 12360
    },
    {
      "entropy": 5.209850692749024,
      "epoch": 1.1878001921229586,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048664214058990546,
      "loss": 4.8698,
      "mean_token_accuracy": 0.22716204673051835,
      "num_tokens": 28348060.0,
      "step": 12365
    },
    {
      "entropy": 5.249106550216675,
      "epoch": 1.1882804995196925,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004866305013056346,
      "loss": 4.9296,
      "mean_token_accuracy": 0.22334240525960922,
      "num_tokens": 28359442.0,
      "step": 12370
    },
    {
      "entropy": 5.296254730224609,
      "epoch": 1.1887608069164266,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004866188571079318,
      "loss": 4.946,
      "mean_token_accuracy": 0.22557690888643264,
      "num_tokens": 28370827.0,
      "step": 12375
    },
    {
      "entropy": 5.372699165344239,
      "epoch": 1.1892411143131605,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004866072079970676,
      "loss": 4.9913,
      "mean_token_accuracy": 0.2106003850698471,
      "num_tokens": 28383115.0,
      "step": 12380
    },
    {
      "entropy": 5.2785381317138675,
      "epoch": 1.1897214217098944,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048659555397331236,
      "loss": 4.8721,
      "mean_token_accuracy": 0.22902147918939592,
      "num_tokens": 28394904.0,
      "step": 12385
    },
    {
      "entropy": 5.232742786407471,
      "epoch": 1.1902017291066282,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004865838950369366,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2200322836637497,
      "num_tokens": 28407357.0,
      "step": 12390
    },
    {
      "entropy": 5.2411457061767575,
      "epoch": 1.1906820365033621,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048657223118821116,
      "loss": 4.8823,
      "mean_token_accuracy": 0.22763336151838304,
      "num_tokens": 28418088.0,
      "step": 12395
    },
    {
      "entropy": 5.138204526901245,
      "epoch": 1.191162343900096,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048656056242740665,
      "loss": 4.7932,
      "mean_token_accuracy": 0.2292204186320305,
      "num_tokens": 28430022.0,
      "step": 12400
    },
    {
      "entropy": 5.264943218231201,
      "epoch": 1.19164265129683,
      "grad_norm": 1.0,
      "learning_rate": 0.0004865488887547942,
      "loss": 4.8876,
      "mean_token_accuracy": 0.22690271884202956,
      "num_tokens": 28440530.0,
      "step": 12405
    },
    {
      "entropy": 5.285682296752929,
      "epoch": 1.192122958693564,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004865372101706446,
      "loss": 4.9728,
      "mean_token_accuracy": 0.22453918009996415,
      "num_tokens": 28452707.0,
      "step": 12410
    },
    {
      "entropy": 5.173147916793823,
      "epoch": 1.1926032660902979,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004865255266752292,
      "loss": 4.7972,
      "mean_token_accuracy": 0.23672958016395568,
      "num_tokens": 28465131.0,
      "step": 12415
    },
    {
      "entropy": 5.19744987487793,
      "epoch": 1.1930835734870318,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004865138382688191,
      "loss": 4.929,
      "mean_token_accuracy": 0.22399083524942398,
      "num_tokens": 28476136.0,
      "step": 12420
    },
    {
      "entropy": 5.254989242553711,
      "epoch": 1.1935638808837656,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004865021449516859,
      "loss": 4.8339,
      "mean_token_accuracy": 0.22820103764533997,
      "num_tokens": 28488374.0,
      "step": 12425
    },
    {
      "entropy": 5.265959024429321,
      "epoch": 1.1940441882804995,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004864904467241008,
      "loss": 4.8937,
      "mean_token_accuracy": 0.22072543650865556,
      "num_tokens": 28499585.0,
      "step": 12430
    },
    {
      "entropy": 5.213364887237549,
      "epoch": 1.1945244956772334,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048647874358633556,
      "loss": 4.833,
      "mean_token_accuracy": 0.23318690061569214,
      "num_tokens": 28510707.0,
      "step": 12435
    },
    {
      "entropy": 5.324516153335571,
      "epoch": 1.1950048030739673,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048646703553866183,
      "loss": 4.9656,
      "mean_token_accuracy": 0.2203390046954155,
      "num_tokens": 28522398.0,
      "step": 12440
    },
    {
      "entropy": 5.23102707862854,
      "epoch": 1.1954851104707012,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004864553225813515,
      "loss": 4.8654,
      "mean_token_accuracy": 0.22697529196739197,
      "num_tokens": 28532949.0,
      "step": 12445
    },
    {
      "entropy": 5.176975393295288,
      "epoch": 1.195965417867435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004864436047146765,
      "loss": 4.8189,
      "mean_token_accuracy": 0.23014866262674333,
      "num_tokens": 28544292.0,
      "step": 12450
    },
    {
      "entropy": 5.288649272918701,
      "epoch": 1.1964457252641691,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048643188193890874,
      "loss": 4.9768,
      "mean_token_accuracy": 0.21843073517084122,
      "num_tokens": 28556430.0,
      "step": 12455
    },
    {
      "entropy": 5.3229070663452145,
      "epoch": 1.196926032660903,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004864201542543206,
      "loss": 4.9529,
      "mean_token_accuracy": 0.22107534855604172,
      "num_tokens": 28568475.0,
      "step": 12460
    },
    {
      "entropy": 5.278416776657105,
      "epoch": 1.197406340057637,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004864084216611843,
      "loss": 4.9158,
      "mean_token_accuracy": 0.21947131156921387,
      "num_tokens": 28579653.0,
      "step": 12465
    },
    {
      "entropy": 5.264760875701905,
      "epoch": 1.1978866474543708,
      "grad_norm": 1.0,
      "learning_rate": 0.00048639668415977207,
      "loss": 4.8603,
      "mean_token_accuracy": 0.22142207026481628,
      "num_tokens": 28590108.0,
      "step": 12470
    },
    {
      "entropy": 5.249595069885254,
      "epoch": 1.1983669548511047,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048638494175035665,
      "loss": 4.8699,
      "mean_token_accuracy": 0.22928204387426376,
      "num_tokens": 28602008.0,
      "step": 12475
    },
    {
      "entropy": 5.205767726898193,
      "epoch": 1.1988472622478386,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004863731944332105,
      "loss": 4.7936,
      "mean_token_accuracy": 0.2310243383049965,
      "num_tokens": 28613286.0,
      "step": 12480
    },
    {
      "entropy": 5.233384132385254,
      "epoch": 1.1993275696445724,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004863614422086065,
      "loss": 4.7947,
      "mean_token_accuracy": 0.23292461037635803,
      "num_tokens": 28625497.0,
      "step": 12485
    },
    {
      "entropy": 5.233258438110352,
      "epoch": 1.1998078770413065,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004863496850768174,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22679537534713745,
      "num_tokens": 28638046.0,
      "step": 12490
    },
    {
      "entropy": 5.1693236351013185,
      "epoch": 1.2002881844380404,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004863379230381162,
      "loss": 4.7974,
      "mean_token_accuracy": 0.2246137261390686,
      "num_tokens": 28647923.0,
      "step": 12495
    },
    {
      "entropy": 5.235925149917603,
      "epoch": 1.2007684918347743,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000486326156092776,
      "loss": 4.8637,
      "mean_token_accuracy": 0.2308898389339447,
      "num_tokens": 28660023.0,
      "step": 12500
    },
    {
      "entropy": 5.235357284545898,
      "epoch": 1.2012487992315082,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048631438424106985,
      "loss": 4.8582,
      "mean_token_accuracy": 0.23112332671880723,
      "num_tokens": 28671568.0,
      "step": 12505
    },
    {
      "entropy": 5.156571960449218,
      "epoch": 1.201729106628242,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048630260748327124,
      "loss": 4.8419,
      "mean_token_accuracy": 0.23185189962387084,
      "num_tokens": 28682897.0,
      "step": 12510
    },
    {
      "entropy": 5.297288656234741,
      "epoch": 1.202209414024976,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048629082581965355,
      "loss": 4.9322,
      "mean_token_accuracy": 0.2247658520936966,
      "num_tokens": 28694067.0,
      "step": 12515
    },
    {
      "entropy": 5.258310604095459,
      "epoch": 1.2026897214217098,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048627903925049033,
      "loss": 4.8578,
      "mean_token_accuracy": 0.22836410701274873,
      "num_tokens": 28705738.0,
      "step": 12520
    },
    {
      "entropy": 5.184260368347168,
      "epoch": 1.2031700288184437,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048626724777605507,
      "loss": 4.8103,
      "mean_token_accuracy": 0.22799164056777954,
      "num_tokens": 28717419.0,
      "step": 12525
    },
    {
      "entropy": 5.30245361328125,
      "epoch": 1.2036503362151778,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004862554513966217,
      "loss": 4.9988,
      "mean_token_accuracy": 0.21163057833909987,
      "num_tokens": 28728587.0,
      "step": 12530
    },
    {
      "entropy": 5.349414587020874,
      "epoch": 1.2041306436119117,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048624365011246405,
      "loss": 4.9764,
      "mean_token_accuracy": 0.22023520469665528,
      "num_tokens": 28740818.0,
      "step": 12535
    },
    {
      "entropy": 5.234213590621948,
      "epoch": 1.2046109510086456,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004862318439238561,
      "loss": 4.8318,
      "mean_token_accuracy": 0.23660051077604294,
      "num_tokens": 28751936.0,
      "step": 12540
    },
    {
      "entropy": 5.235257291793824,
      "epoch": 1.2050912584053795,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000486220032831072,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22376603931188582,
      "num_tokens": 28763271.0,
      "step": 12545
    },
    {
      "entropy": 5.327375793457032,
      "epoch": 1.2055715658021133,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004862082168343859,
      "loss": 4.9312,
      "mean_token_accuracy": 0.22724085599184035,
      "num_tokens": 28774282.0,
      "step": 12550
    },
    {
      "entropy": 5.286591339111328,
      "epoch": 1.2060518731988472,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004861963959340722,
      "loss": 4.9739,
      "mean_token_accuracy": 0.22148910760879517,
      "num_tokens": 28785826.0,
      "step": 12555
    },
    {
      "entropy": 5.18774824142456,
      "epoch": 1.206532180595581,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004861845701304053,
      "loss": 4.801,
      "mean_token_accuracy": 0.2329452157020569,
      "num_tokens": 28797669.0,
      "step": 12560
    },
    {
      "entropy": 5.287655544281006,
      "epoch": 1.2070124879923152,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048617273942365977,
      "loss": 4.9713,
      "mean_token_accuracy": 0.2200448676943779,
      "num_tokens": 28808438.0,
      "step": 12565
    },
    {
      "entropy": 5.2233113765716555,
      "epoch": 1.207492795389049,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004861609038141103,
      "loss": 4.8105,
      "mean_token_accuracy": 0.23219478130340576,
      "num_tokens": 28819707.0,
      "step": 12570
    },
    {
      "entropy": 5.256862020492553,
      "epoch": 1.207973102785783,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048614906330203165,
      "loss": 4.917,
      "mean_token_accuracy": 0.22536155730485916,
      "num_tokens": 28831829.0,
      "step": 12575
    },
    {
      "entropy": 5.204576587677002,
      "epoch": 1.2084534101825168,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004861372178876987,
      "loss": 4.8812,
      "mean_token_accuracy": 0.226816026866436,
      "num_tokens": 28843429.0,
      "step": 12580
    },
    {
      "entropy": 5.207948637008667,
      "epoch": 1.2089337175792507,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048612536757138653,
      "loss": 4.8144,
      "mean_token_accuracy": 0.23279892802238464,
      "num_tokens": 28856239.0,
      "step": 12585
    },
    {
      "entropy": 5.280081701278687,
      "epoch": 1.2094140249759846,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004861135123533702,
      "loss": 4.8981,
      "mean_token_accuracy": 0.23001544028520585,
      "num_tokens": 28868599.0,
      "step": 12590
    },
    {
      "entropy": 5.25154480934143,
      "epoch": 1.2098943323727185,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048610165223392503,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22822480499744416,
      "num_tokens": 28880234.0,
      "step": 12595
    },
    {
      "entropy": 5.305441427230835,
      "epoch": 1.2103746397694524,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004860897872133263,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22299443930387497,
      "num_tokens": 28892198.0,
      "step": 12600
    },
    {
      "entropy": 5.283140373229981,
      "epoch": 1.2108549471661862,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004860779172918496,
      "loss": 4.9834,
      "mean_token_accuracy": 0.21964865028858185,
      "num_tokens": 28904153.0,
      "step": 12605
    },
    {
      "entropy": 5.316070032119751,
      "epoch": 1.2113352545629203,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004860660424697704,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2188755452632904,
      "num_tokens": 28915284.0,
      "step": 12610
    },
    {
      "entropy": 5.267867136001587,
      "epoch": 1.2118155619596542,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048605416274736434,
      "loss": 4.905,
      "mean_token_accuracy": 0.22327355146408082,
      "num_tokens": 28928168.0,
      "step": 12615
    },
    {
      "entropy": 5.215418863296509,
      "epoch": 1.2122958693563881,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048604227812490744,
      "loss": 4.8037,
      "mean_token_accuracy": 0.23381783962249755,
      "num_tokens": 28938548.0,
      "step": 12620
    },
    {
      "entropy": 5.238486862182617,
      "epoch": 1.212776176753122,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048603038860267546,
      "loss": 4.8937,
      "mean_token_accuracy": 0.22736132144927979,
      "num_tokens": 28949252.0,
      "step": 12625
    },
    {
      "entropy": 5.278500175476074,
      "epoch": 1.2132564841498559,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004860184941809445,
      "loss": 4.8979,
      "mean_token_accuracy": 0.2273317828774452,
      "num_tokens": 28960758.0,
      "step": 12630
    },
    {
      "entropy": 5.286111879348755,
      "epoch": 1.2137367915465898,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048600659485999073,
      "loss": 4.8744,
      "mean_token_accuracy": 0.23023439794778824,
      "num_tokens": 28972604.0,
      "step": 12635
    },
    {
      "entropy": 5.22264518737793,
      "epoch": 1.2142170989433236,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048599469064009027,
      "loss": 4.8742,
      "mean_token_accuracy": 0.23308848589658737,
      "num_tokens": 28983617.0,
      "step": 12640
    },
    {
      "entropy": 5.2341522693634035,
      "epoch": 1.2146974063400577,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048598278152151974,
      "loss": 4.8989,
      "mean_token_accuracy": 0.22609585970640184,
      "num_tokens": 28994277.0,
      "step": 12645
    },
    {
      "entropy": 5.333468341827393,
      "epoch": 1.2151777137367916,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004859708675045555,
      "loss": 4.9456,
      "mean_token_accuracy": 0.2273564100265503,
      "num_tokens": 29004844.0,
      "step": 12650
    },
    {
      "entropy": 5.151684188842774,
      "epoch": 1.2156580211335255,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004859589485894741,
      "loss": 4.7991,
      "mean_token_accuracy": 0.23233932256698608,
      "num_tokens": 29015978.0,
      "step": 12655
    },
    {
      "entropy": 5.287998533248901,
      "epoch": 1.2161383285302594,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004859470247765524,
      "loss": 4.9264,
      "mean_token_accuracy": 0.22193968743085862,
      "num_tokens": 29026966.0,
      "step": 12660
    },
    {
      "entropy": 5.277869415283203,
      "epoch": 1.2166186359269933,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004859350960660671,
      "loss": 4.8422,
      "mean_token_accuracy": 0.22963226735591888,
      "num_tokens": 29037943.0,
      "step": 12665
    },
    {
      "entropy": 5.305451965332031,
      "epoch": 1.2170989433237271,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004859231624582953,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2217606395483017,
      "num_tokens": 29049939.0,
      "step": 12670
    },
    {
      "entropy": 5.220596408843994,
      "epoch": 1.217579250720461,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048591122395351394,
      "loss": 4.9269,
      "mean_token_accuracy": 0.22167796790599822,
      "num_tokens": 29062442.0,
      "step": 12675
    },
    {
      "entropy": 5.199642610549927,
      "epoch": 1.218059558117195,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004858992805520003,
      "loss": 4.9015,
      "mean_token_accuracy": 0.22763869613409043,
      "num_tokens": 29074240.0,
      "step": 12680
    },
    {
      "entropy": 5.318103408813476,
      "epoch": 1.218539865513929,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048588733225403153,
      "loss": 4.9241,
      "mean_token_accuracy": 0.21988107711076738,
      "num_tokens": 29085550.0,
      "step": 12685
    },
    {
      "entropy": 5.255595302581787,
      "epoch": 1.219020172910663,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004858753790598851,
      "loss": 4.858,
      "mean_token_accuracy": 0.22498361319303511,
      "num_tokens": 29097092.0,
      "step": 12690
    },
    {
      "entropy": 5.226280736923218,
      "epoch": 1.2195004803073968,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004858634209698386,
      "loss": 4.8975,
      "mean_token_accuracy": 0.228962242603302,
      "num_tokens": 29108119.0,
      "step": 12695
    },
    {
      "entropy": 5.227525949478149,
      "epoch": 1.2199807877041307,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048585145798416956,
      "loss": 4.8804,
      "mean_token_accuracy": 0.22378256916999817,
      "num_tokens": 29120605.0,
      "step": 12700
    },
    {
      "entropy": 5.275301742553711,
      "epoch": 1.2204610951008645,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004858394901031558,
      "loss": 4.8726,
      "mean_token_accuracy": 0.22392308413982392,
      "num_tokens": 29131582.0,
      "step": 12705
    },
    {
      "entropy": 5.3092803955078125,
      "epoch": 1.2209414024975984,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004858275173270751,
      "loss": 4.8979,
      "mean_token_accuracy": 0.21994855999946594,
      "num_tokens": 29143436.0,
      "step": 12710
    },
    {
      "entropy": 5.240077543258667,
      "epoch": 1.2214217098943323,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048581553965620553,
      "loss": 4.8169,
      "mean_token_accuracy": 0.23065620809793472,
      "num_tokens": 29154445.0,
      "step": 12715
    },
    {
      "entropy": 5.265241241455078,
      "epoch": 1.2219020172910664,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048580355709082506,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22355904281139374,
      "num_tokens": 29164599.0,
      "step": 12720
    },
    {
      "entropy": 5.219333982467651,
      "epoch": 1.2223823246878003,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000485791569631212,
      "loss": 4.8895,
      "mean_token_accuracy": 0.2260858491063118,
      "num_tokens": 29176275.0,
      "step": 12725
    },
    {
      "entropy": 5.3178504467010494,
      "epoch": 1.2228626320845342,
      "grad_norm": 0.875,
      "learning_rate": 0.0004857795772776446,
      "loss": 4.8955,
      "mean_token_accuracy": 0.22589892745018006,
      "num_tokens": 29189102.0,
      "step": 12730
    },
    {
      "entropy": 5.285462236404419,
      "epoch": 1.223342939481268,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048576758003040127,
      "loss": 4.9333,
      "mean_token_accuracy": 0.2241787075996399,
      "num_tokens": 29200953.0,
      "step": 12735
    },
    {
      "entropy": 5.268452501296997,
      "epoch": 1.223823246878002,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048575557788976066,
      "loss": 4.9346,
      "mean_token_accuracy": 0.21562522202730178,
      "num_tokens": 29212942.0,
      "step": 12740
    },
    {
      "entropy": 5.212307500839233,
      "epoch": 1.2243035542747358,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004857435708560013,
      "loss": 4.8207,
      "mean_token_accuracy": 0.2306500941514969,
      "num_tokens": 29224949.0,
      "step": 12745
    },
    {
      "entropy": 5.334169149398804,
      "epoch": 1.2247838616714697,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048573155892940204,
      "loss": 4.988,
      "mean_token_accuracy": 0.21955650746822358,
      "num_tokens": 29236044.0,
      "step": 12750
    },
    {
      "entropy": 5.240275621414185,
      "epoch": 1.2252641690682036,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048571954211024164,
      "loss": 4.8831,
      "mean_token_accuracy": 0.22719476968050004,
      "num_tokens": 29248084.0,
      "step": 12755
    },
    {
      "entropy": 5.254735088348388,
      "epoch": 1.2257444764649374,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048570752039879924,
      "loss": 4.7835,
      "mean_token_accuracy": 0.23290664106607437,
      "num_tokens": 29258710.0,
      "step": 12760
    },
    {
      "entropy": 5.322985982894897,
      "epoch": 1.2262247838616716,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004856954937953539,
      "loss": 4.9712,
      "mean_token_accuracy": 0.21359403431415558,
      "num_tokens": 29270173.0,
      "step": 12765
    },
    {
      "entropy": 5.297636985778809,
      "epoch": 1.2267050912584054,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004856834623001848,
      "loss": 4.9604,
      "mean_token_accuracy": 0.2223913624882698,
      "num_tokens": 29280407.0,
      "step": 12770
    },
    {
      "entropy": 5.207039451599121,
      "epoch": 1.2271853986551393,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004856714259135713,
      "loss": 4.8401,
      "mean_token_accuracy": 0.23594015687704087,
      "num_tokens": 29292287.0,
      "step": 12775
    },
    {
      "entropy": 5.261013650894165,
      "epoch": 1.2276657060518732,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004856593846357929,
      "loss": 4.9652,
      "mean_token_accuracy": 0.2202865481376648,
      "num_tokens": 29303099.0,
      "step": 12780
    },
    {
      "entropy": 5.357610893249512,
      "epoch": 1.228146013448607,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004856473384671291,
      "loss": 4.9961,
      "mean_token_accuracy": 0.220786751806736,
      "num_tokens": 29314445.0,
      "step": 12785
    },
    {
      "entropy": 5.212546873092651,
      "epoch": 1.228626320845341,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048563528740785955,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22397034168243407,
      "num_tokens": 29325309.0,
      "step": 12790
    },
    {
      "entropy": 5.223859786987305,
      "epoch": 1.2291066282420748,
      "grad_norm": 0.875,
      "learning_rate": 0.00048562323145826414,
      "loss": 4.8183,
      "mean_token_accuracy": 0.23350467383861542,
      "num_tokens": 29338582.0,
      "step": 12795
    },
    {
      "entropy": 5.266969394683838,
      "epoch": 1.229586935638809,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004856111706186227,
      "loss": 4.8844,
      "mean_token_accuracy": 0.2269684687256813,
      "num_tokens": 29349875.0,
      "step": 12800
    },
    {
      "entropy": 5.345507335662842,
      "epoch": 1.2300672430355428,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048559910488921534,
      "loss": 5.0028,
      "mean_token_accuracy": 0.21599222868680953,
      "num_tokens": 29361800.0,
      "step": 12805
    },
    {
      "entropy": 5.235065841674805,
      "epoch": 1.2305475504322767,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000485587034270322,
      "loss": 4.8754,
      "mean_token_accuracy": 0.22852547764778136,
      "num_tokens": 29372795.0,
      "step": 12810
    },
    {
      "entropy": 5.291785860061646,
      "epoch": 1.2310278578290106,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000485574958762223,
      "loss": 4.9609,
      "mean_token_accuracy": 0.22165000140666963,
      "num_tokens": 29385391.0,
      "step": 12815
    },
    {
      "entropy": 5.203650283813476,
      "epoch": 1.2315081652257445,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048556287836519886,
      "loss": 4.8043,
      "mean_token_accuracy": 0.235040283203125,
      "num_tokens": 29397113.0,
      "step": 12820
    },
    {
      "entropy": 5.343577480316162,
      "epoch": 1.2319884726224783,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004855507930795299,
      "loss": 5.0016,
      "mean_token_accuracy": 0.2221655011177063,
      "num_tokens": 29407552.0,
      "step": 12825
    },
    {
      "entropy": 5.291733503341675,
      "epoch": 1.2324687800192122,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048553870290549665,
      "loss": 4.9176,
      "mean_token_accuracy": 0.22669115662574768,
      "num_tokens": 29418500.0,
      "step": 12830
    },
    {
      "entropy": 5.318113470077515,
      "epoch": 1.232949087415946,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048552660784338,
      "loss": 4.9212,
      "mean_token_accuracy": 0.2243613436818123,
      "num_tokens": 29430335.0,
      "step": 12835
    },
    {
      "entropy": 5.2694066047668455,
      "epoch": 1.23342939481268,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004855145078934606,
      "loss": 4.8989,
      "mean_token_accuracy": 0.22033340036869048,
      "num_tokens": 29441435.0,
      "step": 12840
    },
    {
      "entropy": 5.275525951385498,
      "epoch": 1.233909702209414,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004855024030560195,
      "loss": 4.8924,
      "mean_token_accuracy": 0.22618498802185058,
      "num_tokens": 29453140.0,
      "step": 12845
    },
    {
      "entropy": 5.1970062255859375,
      "epoch": 1.234390009606148,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004854902933313376,
      "loss": 4.8296,
      "mean_token_accuracy": 0.23203448802232743,
      "num_tokens": 29464572.0,
      "step": 12850
    },
    {
      "entropy": 5.319059801101685,
      "epoch": 1.2348703170028819,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048547817871969607,
      "loss": 5.0154,
      "mean_token_accuracy": 0.21419730335474013,
      "num_tokens": 29477069.0,
      "step": 12855
    },
    {
      "entropy": 5.351066780090332,
      "epoch": 1.2353506243996157,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048546605922137633,
      "loss": 4.9135,
      "mean_token_accuracy": 0.22674830704927446,
      "num_tokens": 29486860.0,
      "step": 12860
    },
    {
      "entropy": 5.380636882781983,
      "epoch": 1.2358309317963496,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004854539348366596,
      "loss": 4.9763,
      "mean_token_accuracy": 0.22133456021547318,
      "num_tokens": 29499129.0,
      "step": 12865
    },
    {
      "entropy": 5.190102434158325,
      "epoch": 1.2363112391930835,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004854418055658274,
      "loss": 4.8419,
      "mean_token_accuracy": 0.23210107535123825,
      "num_tokens": 29510764.0,
      "step": 12870
    },
    {
      "entropy": 5.2004303455352785,
      "epoch": 1.2367915465898176,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048542967140916134,
      "loss": 4.9222,
      "mean_token_accuracy": 0.2262816086411476,
      "num_tokens": 29522882.0,
      "step": 12875
    },
    {
      "entropy": 5.311832427978516,
      "epoch": 1.2372718539865515,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004854175323669432,
      "loss": 4.8693,
      "mean_token_accuracy": 0.22457033693790435,
      "num_tokens": 29533348.0,
      "step": 12880
    },
    {
      "entropy": 5.2541898727417,
      "epoch": 1.2377521613832854,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004854053884394547,
      "loss": 4.857,
      "mean_token_accuracy": 0.2307183936238289,
      "num_tokens": 29545649.0,
      "step": 12885
    },
    {
      "entropy": 5.26476149559021,
      "epoch": 1.2382324687800192,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048539323962697796,
      "loss": 4.904,
      "mean_token_accuracy": 0.22181587517261506,
      "num_tokens": 29558252.0,
      "step": 12890
    },
    {
      "entropy": 5.360343980789184,
      "epoch": 1.2387127761767531,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004853810859297949,
      "loss": 5.0112,
      "mean_token_accuracy": 0.22122805267572404,
      "num_tokens": 29569495.0,
      "step": 12895
    },
    {
      "entropy": 5.273426723480225,
      "epoch": 1.239193083573487,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048536892734818773,
      "loss": 4.8996,
      "mean_token_accuracy": 0.22642304748296738,
      "num_tokens": 29582167.0,
      "step": 12900
    },
    {
      "entropy": 5.29416561126709,
      "epoch": 1.239673390970221,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004853567638824387,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2236578106880188,
      "num_tokens": 29593964.0,
      "step": 12905
    },
    {
      "entropy": 5.2863058090209964,
      "epoch": 1.2401536983669548,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048534459553283026,
      "loss": 4.8659,
      "mean_token_accuracy": 0.22707002758979797,
      "num_tokens": 29604844.0,
      "step": 12910
    },
    {
      "entropy": 5.301629495620728,
      "epoch": 1.2406340057636887,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004853324222996449,
      "loss": 4.9183,
      "mean_token_accuracy": 0.22700344324111937,
      "num_tokens": 29615219.0,
      "step": 12915
    },
    {
      "entropy": 5.365880727767944,
      "epoch": 1.2411143131604228,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048532024418316525,
      "loss": 4.9997,
      "mean_token_accuracy": 0.21911056190729142,
      "num_tokens": 29626472.0,
      "step": 12920
    },
    {
      "entropy": 5.234377241134643,
      "epoch": 1.2415946205571566,
      "grad_norm": 1.0,
      "learning_rate": 0.0004853080611836741,
      "loss": 4.8692,
      "mean_token_accuracy": 0.22924379110336304,
      "num_tokens": 29637966.0,
      "step": 12925
    },
    {
      "entropy": 5.212795972824097,
      "epoch": 1.2420749279538905,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048529587330145427,
      "loss": 4.812,
      "mean_token_accuracy": 0.23247532844543456,
      "num_tokens": 29648730.0,
      "step": 12930
    },
    {
      "entropy": 5.201146841049194,
      "epoch": 1.2425552353506244,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048528368053678863,
      "loss": 4.8352,
      "mean_token_accuracy": 0.22391604632139206,
      "num_tokens": 29660576.0,
      "step": 12935
    },
    {
      "entropy": 5.23126449584961,
      "epoch": 1.2430355427473583,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004852714828899604,
      "loss": 4.8614,
      "mean_token_accuracy": 0.2295791760087013,
      "num_tokens": 29672906.0,
      "step": 12940
    },
    {
      "entropy": 5.231838083267212,
      "epoch": 1.2435158501440922,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048525928036125264,
      "loss": 4.8501,
      "mean_token_accuracy": 0.23300151973962785,
      "num_tokens": 29685360.0,
      "step": 12945
    },
    {
      "entropy": 5.260158538818359,
      "epoch": 1.243996157540826,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048524707295094884,
      "loss": 4.8526,
      "mean_token_accuracy": 0.22357651740312576,
      "num_tokens": 29697257.0,
      "step": 12950
    },
    {
      "entropy": 5.214135551452637,
      "epoch": 1.2444764649375601,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004852348606593322,
      "loss": 4.8169,
      "mean_token_accuracy": 0.2303972452878952,
      "num_tokens": 29707877.0,
      "step": 12955
    },
    {
      "entropy": 5.256558275222778,
      "epoch": 1.244956772334294,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048522264348668646,
      "loss": 4.8893,
      "mean_token_accuracy": 0.22467488795518875,
      "num_tokens": 29719358.0,
      "step": 12960
    },
    {
      "entropy": 5.190018272399902,
      "epoch": 1.245437079731028,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004852104214332951,
      "loss": 4.7583,
      "mean_token_accuracy": 0.24494388848543167,
      "num_tokens": 29730383.0,
      "step": 12965
    },
    {
      "entropy": 5.295642900466919,
      "epoch": 1.2459173871277618,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048519819449944205,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22986571192741395,
      "num_tokens": 29741142.0,
      "step": 12970
    },
    {
      "entropy": 5.219013738632202,
      "epoch": 1.2463976945244957,
      "grad_norm": 0.828125,
      "learning_rate": 0.000485185962685411,
      "loss": 4.8577,
      "mean_token_accuracy": 0.22742039412260057,
      "num_tokens": 29754618.0,
      "step": 12975
    },
    {
      "entropy": 5.26318769454956,
      "epoch": 1.2468780019212296,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000485173725991486,
      "loss": 4.8779,
      "mean_token_accuracy": 0.22579114735126496,
      "num_tokens": 29767115.0,
      "step": 12980
    },
    {
      "entropy": 5.321503114700318,
      "epoch": 1.2473583093179634,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048516148441795124,
      "loss": 4.9481,
      "mean_token_accuracy": 0.21918575912714006,
      "num_tokens": 29778165.0,
      "step": 12985
    },
    {
      "entropy": 5.326020336151123,
      "epoch": 1.2478386167146973,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004851492379650908,
      "loss": 5.0062,
      "mean_token_accuracy": 0.2192527174949646,
      "num_tokens": 29790528.0,
      "step": 12990
    },
    {
      "entropy": 5.266337728500366,
      "epoch": 1.2483189241114312,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004851369866331891,
      "loss": 4.8305,
      "mean_token_accuracy": 0.22612589448690415,
      "num_tokens": 29801709.0,
      "step": 12995
    },
    {
      "entropy": 5.264732122421265,
      "epoch": 1.2487992315081653,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004851247304225306,
      "loss": 4.9119,
      "mean_token_accuracy": 0.22029422521591185,
      "num_tokens": 29812963.0,
      "step": 13000
    },
    {
      "entropy": 5.303981065750122,
      "epoch": 1.2492795389048992,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004851124693333997,
      "loss": 4.8745,
      "mean_token_accuracy": 0.22980688661336898,
      "num_tokens": 29823711.0,
      "step": 13005
    },
    {
      "entropy": 5.251740980148315,
      "epoch": 1.249759846301633,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004851002033660812,
      "loss": 4.8501,
      "mean_token_accuracy": 0.23024596273899078,
      "num_tokens": 29834038.0,
      "step": 13010
    },
    {
      "entropy": 5.275460433959961,
      "epoch": 1.250240153698367,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048508793252085994,
      "loss": 4.881,
      "mean_token_accuracy": 0.2286018118262291,
      "num_tokens": 29844759.0,
      "step": 13015
    },
    {
      "entropy": 5.185720300674438,
      "epoch": 1.2507204610951008,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004850756567980206,
      "loss": 4.7561,
      "mean_token_accuracy": 0.2317831814289093,
      "num_tokens": 29855643.0,
      "step": 13020
    },
    {
      "entropy": 5.2829090595245365,
      "epoch": 1.2512007684918347,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048506337619784836,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2240532472729683,
      "num_tokens": 29866917.0,
      "step": 13025
    },
    {
      "entropy": 5.333576917648315,
      "epoch": 1.2516810758885688,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004850510907206283,
      "loss": 5.0287,
      "mean_token_accuracy": 0.2138984888792038,
      "num_tokens": 29878937.0,
      "step": 13030
    },
    {
      "entropy": 5.20754599571228,
      "epoch": 1.2521613832853027,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048503880036664555,
      "loss": 4.8442,
      "mean_token_accuracy": 0.2304176226258278,
      "num_tokens": 29889544.0,
      "step": 13035
    },
    {
      "entropy": 5.213012027740478,
      "epoch": 1.2526416906820366,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004850265051361857,
      "loss": 4.8455,
      "mean_token_accuracy": 0.2272100955247879,
      "num_tokens": 29901919.0,
      "step": 13040
    },
    {
      "entropy": 5.262656545639038,
      "epoch": 1.2531219980787704,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004850142050295339,
      "loss": 4.8967,
      "mean_token_accuracy": 0.2247050881385803,
      "num_tokens": 29913870.0,
      "step": 13045
    },
    {
      "entropy": 5.207606554031372,
      "epoch": 1.2536023054755043,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048500190004697595,
      "loss": 4.8549,
      "mean_token_accuracy": 0.23224006593227386,
      "num_tokens": 29925180.0,
      "step": 13050
    },
    {
      "entropy": 5.156466341018676,
      "epoch": 1.2540826128722382,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004849895901887974,
      "loss": 4.7822,
      "mean_token_accuracy": 0.23526470214128495,
      "num_tokens": 29936433.0,
      "step": 13055
    },
    {
      "entropy": 5.329647541046143,
      "epoch": 1.254562920268972,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004849772754552842,
      "loss": 4.9593,
      "mean_token_accuracy": 0.22186495065689088,
      "num_tokens": 29948891.0,
      "step": 13060
    },
    {
      "entropy": 5.234441089630127,
      "epoch": 1.255043227665706,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048496495584672214,
      "loss": 4.7681,
      "mean_token_accuracy": 0.23815123438835145,
      "num_tokens": 29960113.0,
      "step": 13065
    },
    {
      "entropy": 5.2727912902832035,
      "epoch": 1.2555235350624399,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048495263136339725,
      "loss": 4.9116,
      "mean_token_accuracy": 0.22050365209579467,
      "num_tokens": 29972168.0,
      "step": 13070
    },
    {
      "entropy": 5.245569849014283,
      "epoch": 1.2560038424591737,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004849403020055956,
      "loss": 4.9253,
      "mean_token_accuracy": 0.2239946097135544,
      "num_tokens": 29982996.0,
      "step": 13075
    },
    {
      "entropy": 5.2576793193817135,
      "epoch": 1.2564841498559078,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048492796777360373,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22372111529111863,
      "num_tokens": 29994088.0,
      "step": 13080
    },
    {
      "entropy": 5.344231271743775,
      "epoch": 1.2569644572526417,
      "grad_norm": 1.0,
      "learning_rate": 0.00048491562866770767,
      "loss": 4.9642,
      "mean_token_accuracy": 0.22285789847373963,
      "num_tokens": 30005403.0,
      "step": 13085
    },
    {
      "entropy": 5.273990774154663,
      "epoch": 1.2574447646493756,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048490328468819404,
      "loss": 4.8875,
      "mean_token_accuracy": 0.23558216094970702,
      "num_tokens": 30015961.0,
      "step": 13090
    },
    {
      "entropy": 5.233916807174682,
      "epoch": 1.2579250720461095,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048489093583534945,
      "loss": 4.8253,
      "mean_token_accuracy": 0.23285600543022156,
      "num_tokens": 30026670.0,
      "step": 13095
    },
    {
      "entropy": 5.227122974395752,
      "epoch": 1.2584053794428434,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004848785821094606,
      "loss": 4.8583,
      "mean_token_accuracy": 0.22883858233690263,
      "num_tokens": 30036711.0,
      "step": 13100
    },
    {
      "entropy": 5.261423778533936,
      "epoch": 1.2588856868395775,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004848662235108142,
      "loss": 4.9481,
      "mean_token_accuracy": 0.22179209738969802,
      "num_tokens": 30047587.0,
      "step": 13105
    },
    {
      "entropy": 5.254379177093506,
      "epoch": 1.2593659942363113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004848538600396973,
      "loss": 4.8891,
      "mean_token_accuracy": 0.22707265317440034,
      "num_tokens": 30059348.0,
      "step": 13110
    },
    {
      "entropy": 5.300622367858887,
      "epoch": 1.2598463016330452,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048484149169639694,
      "loss": 4.8975,
      "mean_token_accuracy": 0.2291546121239662,
      "num_tokens": 30070485.0,
      "step": 13115
    },
    {
      "entropy": 5.225955677032471,
      "epoch": 1.260326609029779,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004848291184812003,
      "loss": 4.8252,
      "mean_token_accuracy": 0.23535728305578232,
      "num_tokens": 30081114.0,
      "step": 13120
    },
    {
      "entropy": 5.250436305999756,
      "epoch": 1.260806916426513,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004848167403943945,
      "loss": 4.9313,
      "mean_token_accuracy": 0.2243501752614975,
      "num_tokens": 30092634.0,
      "step": 13125
    },
    {
      "entropy": 5.240738344192505,
      "epoch": 1.2612872238232469,
      "grad_norm": 0.875,
      "learning_rate": 0.00048480435743626703,
      "loss": 4.8753,
      "mean_token_accuracy": 0.2302903950214386,
      "num_tokens": 30104205.0,
      "step": 13130
    },
    {
      "entropy": 5.2762439250946045,
      "epoch": 1.2617675312199808,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004847919696071054,
      "loss": 4.9049,
      "mean_token_accuracy": 0.22426530420780183,
      "num_tokens": 30116978.0,
      "step": 13135
    },
    {
      "entropy": 5.261464357376099,
      "epoch": 1.2622478386167146,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048477957690719716,
      "loss": 4.8084,
      "mean_token_accuracy": 0.23075108379125595,
      "num_tokens": 30128549.0,
      "step": 13140
    },
    {
      "entropy": 5.300140333175659,
      "epoch": 1.2627281460134485,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004847671793368301,
      "loss": 4.9283,
      "mean_token_accuracy": 0.22608596980571746,
      "num_tokens": 30139492.0,
      "step": 13145
    },
    {
      "entropy": 5.255127096176148,
      "epoch": 1.2632084534101824,
      "grad_norm": 0.90625,
      "learning_rate": 0.000484754776896292,
      "loss": 4.8703,
      "mean_token_accuracy": 0.23080528080463408,
      "num_tokens": 30150450.0,
      "step": 13150
    },
    {
      "entropy": 5.289035844802856,
      "epoch": 1.2636887608069165,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004847423695858708,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22550304979085922,
      "num_tokens": 30162204.0,
      "step": 13155
    },
    {
      "entropy": 5.293781518936157,
      "epoch": 1.2641690682036504,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048472995740585456,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22722245901823043,
      "num_tokens": 30172574.0,
      "step": 13160
    },
    {
      "entropy": 5.183301782608032,
      "epoch": 1.2646493756003843,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004847175403565316,
      "loss": 4.7965,
      "mean_token_accuracy": 0.2324737161397934,
      "num_tokens": 30183957.0,
      "step": 13165
    },
    {
      "entropy": 5.259455728530884,
      "epoch": 1.2651296829971181,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048470511843818996,
      "loss": 4.8693,
      "mean_token_accuracy": 0.23278642892837526,
      "num_tokens": 30194207.0,
      "step": 13170
    },
    {
      "entropy": 5.194502449035644,
      "epoch": 1.265609990393852,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004846926916511182,
      "loss": 4.8133,
      "mean_token_accuracy": 0.23785427063703538,
      "num_tokens": 30205180.0,
      "step": 13175
    },
    {
      "entropy": 5.24215612411499,
      "epoch": 1.266090297790586,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004846802599956048,
      "loss": 4.8731,
      "mean_token_accuracy": 0.22989263981580735,
      "num_tokens": 30217734.0,
      "step": 13180
    },
    {
      "entropy": 5.357070398330689,
      "epoch": 1.26657060518732,
      "grad_norm": 1.0,
      "learning_rate": 0.00048466782347193847,
      "loss": 5.0873,
      "mean_token_accuracy": 0.21310788989067078,
      "num_tokens": 30229835.0,
      "step": 13185
    },
    {
      "entropy": 5.319282913208008,
      "epoch": 1.267050912584054,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048465538208040775,
      "loss": 4.9267,
      "mean_token_accuracy": 0.22596296221017836,
      "num_tokens": 30241932.0,
      "step": 13190
    },
    {
      "entropy": 5.323197698593139,
      "epoch": 1.2675312199807878,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048464293582130166,
      "loss": 4.938,
      "mean_token_accuracy": 0.22124491930007933,
      "num_tokens": 30253149.0,
      "step": 13195
    },
    {
      "entropy": 5.27756552696228,
      "epoch": 1.2680115273775217,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004846304846949091,
      "loss": 4.8482,
      "mean_token_accuracy": 0.22936523854732513,
      "num_tokens": 30264083.0,
      "step": 13200
    },
    {
      "entropy": 5.233906698226929,
      "epoch": 1.2684918347742555,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048461802870151916,
      "loss": 4.8196,
      "mean_token_accuracy": 0.2354392945766449,
      "num_tokens": 30274832.0,
      "step": 13205
    },
    {
      "entropy": 5.19139838218689,
      "epoch": 1.2689721421709894,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048460556784142106,
      "loss": 4.8428,
      "mean_token_accuracy": 0.22619154453277587,
      "num_tokens": 30284945.0,
      "step": 13210
    },
    {
      "entropy": 5.318532133102417,
      "epoch": 1.2694524495677233,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048459310211490406,
      "loss": 4.8904,
      "mean_token_accuracy": 0.2321361392736435,
      "num_tokens": 30295133.0,
      "step": 13215
    },
    {
      "entropy": 5.254549837112426,
      "epoch": 1.2699327569644572,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004845806315222576,
      "loss": 4.9357,
      "mean_token_accuracy": 0.2255373015999794,
      "num_tokens": 30305268.0,
      "step": 13220
    },
    {
      "entropy": 5.190180921554566,
      "epoch": 1.270413064361191,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004845681560637711,
      "loss": 4.8185,
      "mean_token_accuracy": 0.22933853566646575,
      "num_tokens": 30317118.0,
      "step": 13225
    },
    {
      "entropy": 5.28556718826294,
      "epoch": 1.270893371757925,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004845556757397344,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22069671154022216,
      "num_tokens": 30328684.0,
      "step": 13230
    },
    {
      "entropy": 5.329628658294678,
      "epoch": 1.271373679154659,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004845431905504372,
      "loss": 4.9895,
      "mean_token_accuracy": 0.21529824435710906,
      "num_tokens": 30341433.0,
      "step": 13235
    },
    {
      "entropy": 5.265971517562866,
      "epoch": 1.271853986551393,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048453070049616926,
      "loss": 4.8462,
      "mean_token_accuracy": 0.23482875078916549,
      "num_tokens": 30353159.0,
      "step": 13240
    },
    {
      "entropy": 5.233815765380859,
      "epoch": 1.2723342939481268,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048451820557722064,
      "loss": 4.9368,
      "mean_token_accuracy": 0.21984797418117524,
      "num_tokens": 30363251.0,
      "step": 13245
    },
    {
      "entropy": 5.356740283966064,
      "epoch": 1.2728146013448607,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004845057057938815,
      "loss": 4.97,
      "mean_token_accuracy": 0.22238073647022247,
      "num_tokens": 30375850.0,
      "step": 13250
    },
    {
      "entropy": 5.335300874710083,
      "epoch": 1.2732949087415946,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048449320114644185,
      "loss": 4.9607,
      "mean_token_accuracy": 0.2206397235393524,
      "num_tokens": 30386839.0,
      "step": 13255
    },
    {
      "entropy": 5.247454881668091,
      "epoch": 1.2737752161383284,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004844806916351922,
      "loss": 4.9484,
      "mean_token_accuracy": 0.22164056301116944,
      "num_tokens": 30398872.0,
      "step": 13260
    },
    {
      "entropy": 5.265853500366211,
      "epoch": 1.2742555235350626,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004844681772604229,
      "loss": 4.8711,
      "mean_token_accuracy": 0.2298185110092163,
      "num_tokens": 30409581.0,
      "step": 13265
    },
    {
      "entropy": 5.276482820510864,
      "epoch": 1.2747358309317964,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048445565802242454,
      "loss": 4.8856,
      "mean_token_accuracy": 0.22777261584997177,
      "num_tokens": 30420209.0,
      "step": 13270
    },
    {
      "entropy": 5.295678234100341,
      "epoch": 1.2752161383285303,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004844431339214878,
      "loss": 4.9219,
      "mean_token_accuracy": 0.2309426873922348,
      "num_tokens": 30432093.0,
      "step": 13275
    },
    {
      "entropy": 5.289564228057861,
      "epoch": 1.2756964457252642,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004844306049579034,
      "loss": 4.853,
      "mean_token_accuracy": 0.2249821349978447,
      "num_tokens": 30442796.0,
      "step": 13280
    },
    {
      "entropy": 5.259756278991699,
      "epoch": 1.276176753121998,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048441807113196216,
      "loss": 4.8781,
      "mean_token_accuracy": 0.2224330946803093,
      "num_tokens": 30455226.0,
      "step": 13285
    },
    {
      "entropy": 5.319718647003174,
      "epoch": 1.276657060518732,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048440553244395517,
      "loss": 4.9666,
      "mean_token_accuracy": 0.22164156287908554,
      "num_tokens": 30467082.0,
      "step": 13290
    },
    {
      "entropy": 5.3239240646362305,
      "epoch": 1.2771373679154658,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048439298889417357,
      "loss": 4.8904,
      "mean_token_accuracy": 0.22027028053998948,
      "num_tokens": 30479051.0,
      "step": 13295
    },
    {
      "entropy": 5.3115763664245605,
      "epoch": 1.2776176753121997,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048438044048290847,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22939268499612808,
      "num_tokens": 30489989.0,
      "step": 13300
    },
    {
      "entropy": 5.285744142532349,
      "epoch": 1.2780979827089336,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048436788721045135,
      "loss": 4.8373,
      "mean_token_accuracy": 0.22701658308506012,
      "num_tokens": 30501533.0,
      "step": 13305
    },
    {
      "entropy": 5.275807809829712,
      "epoch": 1.2785782901056677,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004843553290770935,
      "loss": 4.9266,
      "mean_token_accuracy": 0.22519933730363845,
      "num_tokens": 30512231.0,
      "step": 13310
    },
    {
      "entropy": 5.244143915176392,
      "epoch": 1.2790585975024016,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004843427660831266,
      "loss": 4.9032,
      "mean_token_accuracy": 0.22736574858427047,
      "num_tokens": 30523204.0,
      "step": 13315
    },
    {
      "entropy": 5.349837160110473,
      "epoch": 1.2795389048991355,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048433019822884235,
      "loss": 5.0142,
      "mean_token_accuracy": 0.21474944651126862,
      "num_tokens": 30534956.0,
      "step": 13320
    },
    {
      "entropy": 5.245494651794433,
      "epoch": 1.2800192122958693,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004843176255145325,
      "loss": 4.8578,
      "mean_token_accuracy": 0.23188574165105819,
      "num_tokens": 30545938.0,
      "step": 13325
    },
    {
      "entropy": 5.26036810874939,
      "epoch": 1.2804995196926032,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004843050479404888,
      "loss": 4.817,
      "mean_token_accuracy": 0.23390018343925476,
      "num_tokens": 30557323.0,
      "step": 13330
    },
    {
      "entropy": 5.2840629577636715,
      "epoch": 1.280979827089337,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048429246550700343,
      "loss": 4.8754,
      "mean_token_accuracy": 0.2290678322315216,
      "num_tokens": 30569960.0,
      "step": 13335
    },
    {
      "entropy": 5.236600255966186,
      "epoch": 1.2814601344860712,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004842798782143686,
      "loss": 4.9157,
      "mean_token_accuracy": 0.22605344355106355,
      "num_tokens": 30581904.0,
      "step": 13340
    },
    {
      "entropy": 5.2965466499328615,
      "epoch": 1.281940441882805,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048426728606287627,
      "loss": 4.8918,
      "mean_token_accuracy": 0.22702774703502654,
      "num_tokens": 30592955.0,
      "step": 13345
    },
    {
      "entropy": 5.303103876113892,
      "epoch": 1.282420749279539,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004842546890528191,
      "loss": 4.9237,
      "mean_token_accuracy": 0.22461764514446259,
      "num_tokens": 30604020.0,
      "step": 13350
    },
    {
      "entropy": 5.25487904548645,
      "epoch": 1.2829010566762729,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004842420871844893,
      "loss": 4.8619,
      "mean_token_accuracy": 0.22306262254714965,
      "num_tokens": 30615623.0,
      "step": 13355
    },
    {
      "entropy": 5.327361822128296,
      "epoch": 1.2833813640730067,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004842294804581796,
      "loss": 4.9186,
      "mean_token_accuracy": 0.2263682872056961,
      "num_tokens": 30626258.0,
      "step": 13360
    },
    {
      "entropy": 5.310267400741577,
      "epoch": 1.2838616714697406,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048421686887418266,
      "loss": 4.8991,
      "mean_token_accuracy": 0.21939824670553207,
      "num_tokens": 30637861.0,
      "step": 13365
    },
    {
      "entropy": 5.308465242385864,
      "epoch": 1.2843419788664745,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004842042524327912,
      "loss": 4.9193,
      "mean_token_accuracy": 0.22672958374023439,
      "num_tokens": 30648835.0,
      "step": 13370
    },
    {
      "entropy": 5.362324953079224,
      "epoch": 1.2848222862632084,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004841916311342983,
      "loss": 4.972,
      "mean_token_accuracy": 0.21618115901947021,
      "num_tokens": 30659117.0,
      "step": 13375
    },
    {
      "entropy": 5.278029251098633,
      "epoch": 1.2853025936599423,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004841790049789969,
      "loss": 4.8836,
      "mean_token_accuracy": 0.23132814913988115,
      "num_tokens": 30670282.0,
      "step": 13380
    },
    {
      "entropy": 5.254837226867676,
      "epoch": 1.2857829010566761,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048416637396718004,
      "loss": 4.9256,
      "mean_token_accuracy": 0.21957337111234665,
      "num_tokens": 30681967.0,
      "step": 13385
    },
    {
      "entropy": 5.343366575241089,
      "epoch": 1.2862632084534102,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004841537380991411,
      "loss": 4.9039,
      "mean_token_accuracy": 0.22540718019008638,
      "num_tokens": 30692803.0,
      "step": 13390
    },
    {
      "entropy": 5.2557862281799315,
      "epoch": 1.2867435158501441,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048414109737517346,
      "loss": 4.7969,
      "mean_token_accuracy": 0.23109558820724488,
      "num_tokens": 30704097.0,
      "step": 13395
    },
    {
      "entropy": 5.295741128921509,
      "epoch": 1.287223823246878,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004841284517955706,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21689423471689223,
      "num_tokens": 30717983.0,
      "step": 13400
    },
    {
      "entropy": 5.177179193496704,
      "epoch": 1.2877041306436119,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000484115801360626,
      "loss": 4.754,
      "mean_token_accuracy": 0.2401679649949074,
      "num_tokens": 30728446.0,
      "step": 13405
    },
    {
      "entropy": 5.149666261672974,
      "epoch": 1.2881844380403458,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004841031460706335,
      "loss": 4.7187,
      "mean_token_accuracy": 0.23736633956432343,
      "num_tokens": 30739587.0,
      "step": 13410
    },
    {
      "entropy": 5.222867679595947,
      "epoch": 1.2886647454370797,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048409048592588683,
      "loss": 4.859,
      "mean_token_accuracy": 0.22946472615003585,
      "num_tokens": 30750093.0,
      "step": 13415
    },
    {
      "entropy": 5.276018810272217,
      "epoch": 1.2891450528338138,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004840778209266799,
      "loss": 4.8852,
      "mean_token_accuracy": 0.2298538327217102,
      "num_tokens": 30761692.0,
      "step": 13420
    },
    {
      "entropy": 5.208982467651367,
      "epoch": 1.2896253602305476,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048406515107330685,
      "loss": 4.8182,
      "mean_token_accuracy": 0.23108222633600234,
      "num_tokens": 30773474.0,
      "step": 13425
    },
    {
      "entropy": 5.381736183166504,
      "epoch": 1.2901056676272815,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048405247636606173,
      "loss": 4.9881,
      "mean_token_accuracy": 0.21474856585264207,
      "num_tokens": 30785464.0,
      "step": 13430
    },
    {
      "entropy": 5.315547370910645,
      "epoch": 1.2905859750240154,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048403979680523894,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22639713287353516,
      "num_tokens": 30796343.0,
      "step": 13435
    },
    {
      "entropy": 5.267082452774048,
      "epoch": 1.2910662824207493,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004840271123911328,
      "loss": 4.8709,
      "mean_token_accuracy": 0.2275972470641136,
      "num_tokens": 30807795.0,
      "step": 13440
    },
    {
      "entropy": 5.241315412521362,
      "epoch": 1.2915465898174832,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004840144231240377,
      "loss": 4.876,
      "mean_token_accuracy": 0.22559971511363983,
      "num_tokens": 30819629.0,
      "step": 13445
    },
    {
      "entropy": 5.245262145996094,
      "epoch": 1.292026897214217,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004840017290042484,
      "loss": 4.8382,
      "mean_token_accuracy": 0.2305685967206955,
      "num_tokens": 30831545.0,
      "step": 13450
    },
    {
      "entropy": 5.284837341308593,
      "epoch": 1.292507204610951,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048398903003205957,
      "loss": 4.8795,
      "mean_token_accuracy": 0.2300468310713768,
      "num_tokens": 30843614.0,
      "step": 13455
    },
    {
      "entropy": 5.2079719543457035,
      "epoch": 1.2929875120076848,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048397632620776604,
      "loss": 4.7973,
      "mean_token_accuracy": 0.23166592866182328,
      "num_tokens": 30853749.0,
      "step": 13460
    },
    {
      "entropy": 5.225417804718018,
      "epoch": 1.293467819404419,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048396361753166276,
      "loss": 4.8235,
      "mean_token_accuracy": 0.22897321581840516,
      "num_tokens": 30865065.0,
      "step": 13465
    },
    {
      "entropy": 5.282008266448974,
      "epoch": 1.2939481268011528,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048395090400404466,
      "loss": 4.9128,
      "mean_token_accuracy": 0.22134197056293486,
      "num_tokens": 30876746.0,
      "step": 13470
    },
    {
      "entropy": 5.3236747741699215,
      "epoch": 1.2944284341978867,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048393818562520715,
      "loss": 4.9435,
      "mean_token_accuracy": 0.22507452517747878,
      "num_tokens": 30889543.0,
      "step": 13475
    },
    {
      "entropy": 5.282744932174682,
      "epoch": 1.2949087415946205,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048392546239544535,
      "loss": 4.9142,
      "mean_token_accuracy": 0.22687785774469377,
      "num_tokens": 30901949.0,
      "step": 13480
    },
    {
      "entropy": 5.296532058715821,
      "epoch": 1.2953890489913544,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004839127343150547,
      "loss": 4.9145,
      "mean_token_accuracy": 0.22464920282363893,
      "num_tokens": 30912660.0,
      "step": 13485
    },
    {
      "entropy": 5.248243761062622,
      "epoch": 1.2958693563880883,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004839000013843307,
      "loss": 4.8892,
      "mean_token_accuracy": 0.2285814791917801,
      "num_tokens": 30924276.0,
      "step": 13490
    },
    {
      "entropy": 5.229885053634644,
      "epoch": 1.2963496637848224,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048388726360356894,
      "loss": 4.822,
      "mean_token_accuracy": 0.2327438399195671,
      "num_tokens": 30935169.0,
      "step": 13495
    },
    {
      "entropy": 5.352291059494019,
      "epoch": 1.2968299711815563,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004838745209730653,
      "loss": 4.9849,
      "mean_token_accuracy": 0.21989069730043412,
      "num_tokens": 30946116.0,
      "step": 13500
    },
    {
      "entropy": 5.314806175231934,
      "epoch": 1.2973102785782902,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048386177349311535,
      "loss": 4.9169,
      "mean_token_accuracy": 0.22850501239299775,
      "num_tokens": 30958509.0,
      "step": 13505
    },
    {
      "entropy": 5.318422937393189,
      "epoch": 1.297790585975024,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004838490211640153,
      "loss": 4.9154,
      "mean_token_accuracy": 0.227701897919178,
      "num_tokens": 30969501.0,
      "step": 13510
    },
    {
      "entropy": 5.2290349960327145,
      "epoch": 1.298270893371758,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004838362639860611,
      "loss": 4.8481,
      "mean_token_accuracy": 0.22880804240703584,
      "num_tokens": 30979768.0,
      "step": 13515
    },
    {
      "entropy": 5.17601318359375,
      "epoch": 1.2987512007684918,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000483823501959549,
      "loss": 4.8351,
      "mean_token_accuracy": 0.23402352929115294,
      "num_tokens": 30992020.0,
      "step": 13520
    },
    {
      "entropy": 5.355890321731567,
      "epoch": 1.2992315081652257,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048381073508477527,
      "loss": 4.9569,
      "mean_token_accuracy": 0.22144169807434083,
      "num_tokens": 31004081.0,
      "step": 13525
    },
    {
      "entropy": 5.374733781814575,
      "epoch": 1.2997118155619596,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048379796336203625,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21453364342451095,
      "num_tokens": 31015776.0,
      "step": 13530
    },
    {
      "entropy": 5.258734464645386,
      "epoch": 1.3001921229586935,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004837851867916286,
      "loss": 4.8106,
      "mean_token_accuracy": 0.2280153974890709,
      "num_tokens": 31028066.0,
      "step": 13535
    },
    {
      "entropy": 5.218272352218628,
      "epoch": 1.3006724303554273,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004837724053738488,
      "loss": 4.862,
      "mean_token_accuracy": 0.23139288723468782,
      "num_tokens": 31040234.0,
      "step": 13540
    },
    {
      "entropy": 5.282900714874268,
      "epoch": 1.3011527377521614,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048375961910899373,
      "loss": 4.9202,
      "mean_token_accuracy": 0.22406985610723495,
      "num_tokens": 31051158.0,
      "step": 13545
    },
    {
      "entropy": 5.336216783523559,
      "epoch": 1.3016330451488953,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004837468279973602,
      "loss": 4.9078,
      "mean_token_accuracy": 0.22586057484149932,
      "num_tokens": 31063404.0,
      "step": 13550
    },
    {
      "entropy": 5.290809106826782,
      "epoch": 1.3021133525456292,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004837340320392451,
      "loss": 4.8754,
      "mean_token_accuracy": 0.22747805267572402,
      "num_tokens": 31073845.0,
      "step": 13555
    },
    {
      "entropy": 5.229689025878907,
      "epoch": 1.302593659942363,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048372123123494563,
      "loss": 4.9002,
      "mean_token_accuracy": 0.2201330617070198,
      "num_tokens": 31086914.0,
      "step": 13560
    },
    {
      "entropy": 5.269000673294068,
      "epoch": 1.303073967339097,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000483708425584759,
      "loss": 4.8311,
      "mean_token_accuracy": 0.23600212037563323,
      "num_tokens": 31098627.0,
      "step": 13565
    },
    {
      "entropy": 5.376222276687622,
      "epoch": 1.3035542747358309,
      "grad_norm": 1.0,
      "learning_rate": 0.0004836956150889825,
      "loss": 5.0804,
      "mean_token_accuracy": 0.21311208307743074,
      "num_tokens": 31110064.0,
      "step": 13570
    },
    {
      "entropy": 5.276593017578125,
      "epoch": 1.304034582132565,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004836827997479134,
      "loss": 4.8311,
      "mean_token_accuracy": 0.2391951322555542,
      "num_tokens": 31121572.0,
      "step": 13575
    },
    {
      "entropy": 5.348249197006226,
      "epoch": 1.3045148895292988,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004836699795618494,
      "loss": 4.9571,
      "mean_token_accuracy": 0.22306533008813859,
      "num_tokens": 31133276.0,
      "step": 13580
    },
    {
      "entropy": 5.331839275360108,
      "epoch": 1.3049951969260327,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004836571545310881,
      "loss": 4.9579,
      "mean_token_accuracy": 0.22275590300559997,
      "num_tokens": 31144524.0,
      "step": 13585
    },
    {
      "entropy": 5.255473518371582,
      "epoch": 1.3054755043227666,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048364432465592723,
      "loss": 4.8384,
      "mean_token_accuracy": 0.23205744177103044,
      "num_tokens": 31155775.0,
      "step": 13590
    },
    {
      "entropy": 5.264395332336425,
      "epoch": 1.3059558117195005,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004836314899366647,
      "loss": 4.813,
      "mean_token_accuracy": 0.23266176581382753,
      "num_tokens": 31166176.0,
      "step": 13595
    },
    {
      "entropy": 5.276184368133545,
      "epoch": 1.3064361191162344,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048361865037359846,
      "loss": 4.8586,
      "mean_token_accuracy": 0.23080765902996064,
      "num_tokens": 31178568.0,
      "step": 13600
    },
    {
      "entropy": 5.331236028671265,
      "epoch": 1.3069164265129682,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048360580596702664,
      "loss": 4.947,
      "mean_token_accuracy": 0.22198917716741562,
      "num_tokens": 31188897.0,
      "step": 13605
    },
    {
      "entropy": 5.269009256362915,
      "epoch": 1.3073967339097021,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048359295671724744,
      "loss": 4.9054,
      "mean_token_accuracy": 0.22687793225049974,
      "num_tokens": 31201370.0,
      "step": 13610
    },
    {
      "entropy": 5.227370691299439,
      "epoch": 1.307877041306436,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004835801026245592,
      "loss": 4.9663,
      "mean_token_accuracy": 0.2248321920633316,
      "num_tokens": 31212704.0,
      "step": 13615
    },
    {
      "entropy": 5.2829649448394775,
      "epoch": 1.30835734870317,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048356724368926035,
      "loss": 4.9241,
      "mean_token_accuracy": 0.22813218981027603,
      "num_tokens": 31224447.0,
      "step": 13620
    },
    {
      "entropy": 5.367801713943481,
      "epoch": 1.308837656099904,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048355437991164937,
      "loss": 4.96,
      "mean_token_accuracy": 0.22481322139501572,
      "num_tokens": 31235648.0,
      "step": 13625
    },
    {
      "entropy": 5.230207538604736,
      "epoch": 1.3093179634966379,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000483541511292025,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22295650094747543,
      "num_tokens": 31248330.0,
      "step": 13630
    },
    {
      "entropy": 5.271143198013306,
      "epoch": 1.3097982708933718,
      "grad_norm": 0.875,
      "learning_rate": 0.00048352863783068594,
      "loss": 4.9694,
      "mean_token_accuracy": 0.2220692440867424,
      "num_tokens": 31260036.0,
      "step": 13635
    },
    {
      "entropy": 5.207055234909058,
      "epoch": 1.3102785782901056,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048351575952793117,
      "loss": 4.7473,
      "mean_token_accuracy": 0.2343164250254631,
      "num_tokens": 31271473.0,
      "step": 13640
    },
    {
      "entropy": 5.189319610595703,
      "epoch": 1.3107588856868395,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004835028763840595,
      "loss": 4.7571,
      "mean_token_accuracy": 0.23460074216127397,
      "num_tokens": 31282045.0,
      "step": 13645
    },
    {
      "entropy": 5.279329490661621,
      "epoch": 1.3112391930835736,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004834899883993703,
      "loss": 5.0041,
      "mean_token_accuracy": 0.22016316652297974,
      "num_tokens": 31292644.0,
      "step": 13650
    },
    {
      "entropy": 5.262202501296997,
      "epoch": 1.3117195004803075,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048347709557416263,
      "loss": 4.9239,
      "mean_token_accuracy": 0.23293898552656173,
      "num_tokens": 31304422.0,
      "step": 13655
    },
    {
      "entropy": 5.222819709777832,
      "epoch": 1.3121998078770414,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004834641979087359,
      "loss": 4.8428,
      "mean_token_accuracy": 0.2261192962527275,
      "num_tokens": 31314845.0,
      "step": 13660
    },
    {
      "entropy": 5.373266410827637,
      "epoch": 1.3126801152737753,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004834512954033894,
      "loss": 4.9732,
      "mean_token_accuracy": 0.22480185478925704,
      "num_tokens": 31325669.0,
      "step": 13665
    },
    {
      "entropy": 5.307702493667603,
      "epoch": 1.3131604226705091,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048343838805842284,
      "loss": 4.8702,
      "mean_token_accuracy": 0.23487961441278457,
      "num_tokens": 31336023.0,
      "step": 13670
    },
    {
      "entropy": 5.290696716308593,
      "epoch": 1.313640730067243,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048342547587413583,
      "loss": 4.955,
      "mean_token_accuracy": 0.22119988203048707,
      "num_tokens": 31347146.0,
      "step": 13675
    },
    {
      "entropy": 5.295088481903076,
      "epoch": 1.314121037463977,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004834125588508282,
      "loss": 4.9417,
      "mean_token_accuracy": 0.2249438941478729,
      "num_tokens": 31357616.0,
      "step": 13680
    },
    {
      "entropy": 5.350727844238281,
      "epoch": 1.3146013448607108,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004833996369887998,
      "loss": 4.9469,
      "mean_token_accuracy": 0.22669768780469896,
      "num_tokens": 31369189.0,
      "step": 13685
    },
    {
      "entropy": 5.2566052913665775,
      "epoch": 1.3150816522574447,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048338671028835063,
      "loss": 4.8168,
      "mean_token_accuracy": 0.23387177884578705,
      "num_tokens": 31380841.0,
      "step": 13690
    },
    {
      "entropy": 5.200211000442505,
      "epoch": 1.3155619596541785,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048337377874978086,
      "loss": 4.8351,
      "mean_token_accuracy": 0.23235509991645814,
      "num_tokens": 31393092.0,
      "step": 13695
    },
    {
      "entropy": 5.2632965564727785,
      "epoch": 1.3160422670509127,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048336084237339067,
      "loss": 4.8902,
      "mean_token_accuracy": 0.22927022874355316,
      "num_tokens": 31404228.0,
      "step": 13700
    },
    {
      "entropy": 5.302024269104004,
      "epoch": 1.3165225744476465,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004833479011594804,
      "loss": 4.9964,
      "mean_token_accuracy": 0.2183295339345932,
      "num_tokens": 31416233.0,
      "step": 13705
    },
    {
      "entropy": 5.335728979110717,
      "epoch": 1.3170028818443804,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048333495510835057,
      "loss": 4.935,
      "mean_token_accuracy": 0.22184885144233704,
      "num_tokens": 31427787.0,
      "step": 13710
    },
    {
      "entropy": 5.156178331375122,
      "epoch": 1.3174831892411143,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048332200422030163,
      "loss": 4.7492,
      "mean_token_accuracy": 0.23013928532600403,
      "num_tokens": 31438736.0,
      "step": 13715
    },
    {
      "entropy": 5.153148126602173,
      "epoch": 1.3179634966378482,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004833090484956345,
      "loss": 4.7669,
      "mean_token_accuracy": 0.23113839626312255,
      "num_tokens": 31449225.0,
      "step": 13720
    },
    {
      "entropy": 5.347294664382934,
      "epoch": 1.318443804034582,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048329608793464966,
      "loss": 4.9684,
      "mean_token_accuracy": 0.22664819806814193,
      "num_tokens": 31461223.0,
      "step": 13725
    },
    {
      "entropy": 5.386128282546997,
      "epoch": 1.3189241114313162,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004832831225376482,
      "loss": 4.9912,
      "mean_token_accuracy": 0.22471692860126496,
      "num_tokens": 31472397.0,
      "step": 13730
    },
    {
      "entropy": 5.211466073989868,
      "epoch": 1.31940441882805,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004832701523049312,
      "loss": 4.8151,
      "mean_token_accuracy": 0.23714623749256133,
      "num_tokens": 31484079.0,
      "step": 13735
    },
    {
      "entropy": 5.231664991378784,
      "epoch": 1.319884726224784,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004832571772367997,
      "loss": 4.8315,
      "mean_token_accuracy": 0.22713046818971633,
      "num_tokens": 31496732.0,
      "step": 13740
    },
    {
      "entropy": 5.291873741149902,
      "epoch": 1.3203650336215178,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048324419733355485,
      "loss": 4.8044,
      "mean_token_accuracy": 0.23548634499311447,
      "num_tokens": 31507601.0,
      "step": 13745
    },
    {
      "entropy": 5.239410734176635,
      "epoch": 1.3208453410182517,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048323121259549805,
      "loss": 4.8429,
      "mean_token_accuracy": 0.23069666028022767,
      "num_tokens": 31519551.0,
      "step": 13750
    },
    {
      "entropy": 5.241321849822998,
      "epoch": 1.3213256484149856,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048321822302293095,
      "loss": 4.921,
      "mean_token_accuracy": 0.22820497155189515,
      "num_tokens": 31531531.0,
      "step": 13755
    },
    {
      "entropy": 5.288639068603516,
      "epoch": 1.3218059558117194,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004832052286161549,
      "loss": 4.8518,
      "mean_token_accuracy": 0.2256810486316681,
      "num_tokens": 31542386.0,
      "step": 13760
    },
    {
      "entropy": 5.283427762985229,
      "epoch": 1.3222862632084533,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048319222937547176,
      "loss": 4.8617,
      "mean_token_accuracy": 0.23201979398727418,
      "num_tokens": 31554142.0,
      "step": 13765
    },
    {
      "entropy": 5.2688703536987305,
      "epoch": 1.3227665706051872,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048317922530118323,
      "loss": 4.901,
      "mean_token_accuracy": 0.22078386098146438,
      "num_tokens": 31566687.0,
      "step": 13770
    },
    {
      "entropy": 5.292323064804077,
      "epoch": 1.3232468780019213,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004831662163935912,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21797942072153093,
      "num_tokens": 31577893.0,
      "step": 13775
    },
    {
      "entropy": 5.399905061721801,
      "epoch": 1.3237271853986552,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048315320265299784,
      "loss": 5.0259,
      "mean_token_accuracy": 0.2186278745532036,
      "num_tokens": 31590483.0,
      "step": 13780
    },
    {
      "entropy": 5.335682201385498,
      "epoch": 1.324207492795389,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048314018407970516,
      "loss": 4.8956,
      "mean_token_accuracy": 0.22432043701410292,
      "num_tokens": 31601373.0,
      "step": 13785
    },
    {
      "entropy": 5.259761381149292,
      "epoch": 1.324687800192123,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048312716067401535,
      "loss": 4.9359,
      "mean_token_accuracy": 0.2184838131070137,
      "num_tokens": 31613981.0,
      "step": 13790
    },
    {
      "entropy": 5.33364315032959,
      "epoch": 1.3251681075888568,
      "grad_norm": 0.96875,
      "learning_rate": 0.000483114132436231,
      "loss": 4.9231,
      "mean_token_accuracy": 0.21579712331295015,
      "num_tokens": 31626727.0,
      "step": 13795
    },
    {
      "entropy": 5.232037496566773,
      "epoch": 1.3256484149855907,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004831010993666545,
      "loss": 4.8069,
      "mean_token_accuracy": 0.23097458630800247,
      "num_tokens": 31638519.0,
      "step": 13800
    },
    {
      "entropy": 5.212914323806762,
      "epoch": 1.3261287223823248,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004830880614655884,
      "loss": 4.8199,
      "mean_token_accuracy": 0.2266894742846489,
      "num_tokens": 31649184.0,
      "step": 13805
    },
    {
      "entropy": 5.274742603302002,
      "epoch": 1.3266090297790587,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048307501873333527,
      "loss": 4.9079,
      "mean_token_accuracy": 0.22649470567703248,
      "num_tokens": 31660266.0,
      "step": 13810
    },
    {
      "entropy": 5.2415228366851805,
      "epoch": 1.3270893371757926,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004830619711701982,
      "loss": 4.8672,
      "mean_token_accuracy": 0.22898614704608916,
      "num_tokens": 31670841.0,
      "step": 13815
    },
    {
      "entropy": 5.252540493011475,
      "epoch": 1.3275696445725265,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004830489187764799,
      "loss": 4.8585,
      "mean_token_accuracy": 0.2272809624671936,
      "num_tokens": 31680786.0,
      "step": 13820
    },
    {
      "entropy": 5.192296743392944,
      "epoch": 1.3280499519692603,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004830358615524835,
      "loss": 4.8626,
      "mean_token_accuracy": 0.22330822050571442,
      "num_tokens": 31692639.0,
      "step": 13825
    },
    {
      "entropy": 5.379722547531128,
      "epoch": 1.3285302593659942,
      "grad_norm": 1.125,
      "learning_rate": 0.00048302279949851215,
      "loss": 4.9659,
      "mean_token_accuracy": 0.22365449517965316,
      "num_tokens": 31703520.0,
      "step": 13830
    },
    {
      "entropy": 5.287240886688233,
      "epoch": 1.329010566762728,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048300973261486906,
      "loss": 4.8232,
      "mean_token_accuracy": 0.23896358907222748,
      "num_tokens": 31715659.0,
      "step": 13835
    },
    {
      "entropy": 5.2732020854949955,
      "epoch": 1.329490874159462,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004829966609018577,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22275891304016113,
      "num_tokens": 31727896.0,
      "step": 13840
    },
    {
      "entropy": 5.253575801849365,
      "epoch": 1.3299711815561959,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048298358435978146,
      "loss": 4.8567,
      "mean_token_accuracy": 0.22768504321575164,
      "num_tokens": 31738618.0,
      "step": 13845
    },
    {
      "entropy": 5.31207275390625,
      "epoch": 1.3304514889529298,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048297050298894394,
      "loss": 4.9308,
      "mean_token_accuracy": 0.22154612243175506,
      "num_tokens": 31750306.0,
      "step": 13850
    },
    {
      "entropy": 5.359532499313355,
      "epoch": 1.3309317963496639,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004829574167896489,
      "loss": 4.9789,
      "mean_token_accuracy": 0.22049502581357955,
      "num_tokens": 31762015.0,
      "step": 13855
    },
    {
      "entropy": 5.239928913116455,
      "epoch": 1.3314121037463977,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048294432576220027,
      "loss": 4.8727,
      "mean_token_accuracy": 0.23183965682983398,
      "num_tokens": 31774120.0,
      "step": 13860
    },
    {
      "entropy": 5.234589385986328,
      "epoch": 1.3318924111431316,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004829312299069017,
      "loss": 4.819,
      "mean_token_accuracy": 0.2354026824235916,
      "num_tokens": 31785485.0,
      "step": 13865
    },
    {
      "entropy": 5.270302486419678,
      "epoch": 1.3323727185398655,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048291812922405755,
      "loss": 4.9181,
      "mean_token_accuracy": 0.22479526996612548,
      "num_tokens": 31795621.0,
      "step": 13870
    },
    {
      "entropy": 5.176194190979004,
      "epoch": 1.3328530259365994,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004829050237139717,
      "loss": 4.8065,
      "mean_token_accuracy": 0.23437039107084273,
      "num_tokens": 31806761.0,
      "step": 13875
    },
    {
      "entropy": 5.32261209487915,
      "epoch": 1.3333333333333333,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004828919133769486,
      "loss": 4.8997,
      "mean_token_accuracy": 0.22242701947689056,
      "num_tokens": 31817996.0,
      "step": 13880
    },
    {
      "entropy": 5.331638193130493,
      "epoch": 1.3338136407300674,
      "grad_norm": 1.0,
      "learning_rate": 0.0004828787982132926,
      "loss": 4.9589,
      "mean_token_accuracy": 0.22055888622999192,
      "num_tokens": 31829888.0,
      "step": 13885
    },
    {
      "entropy": 5.325413513183594,
      "epoch": 1.3342939481268012,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048286567822330815,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22950852066278457,
      "num_tokens": 31842582.0,
      "step": 13890
    },
    {
      "entropy": 5.205611371994019,
      "epoch": 1.3347742555235351,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004828525534072999,
      "loss": 4.7906,
      "mean_token_accuracy": 0.2312016025185585,
      "num_tokens": 31854765.0,
      "step": 13895
    },
    {
      "entropy": 5.26736273765564,
      "epoch": 1.335254562920269,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048283942376557254,
      "loss": 4.8583,
      "mean_token_accuracy": 0.22360893934965134,
      "num_tokens": 31865505.0,
      "step": 13900
    },
    {
      "entropy": 5.212670469284058,
      "epoch": 1.3357348703170029,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048282628929843097,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2274489313364029,
      "num_tokens": 31876599.0,
      "step": 13905
    },
    {
      "entropy": 5.299495601654053,
      "epoch": 1.3362151777137368,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048281315000617996,
      "loss": 4.857,
      "mean_token_accuracy": 0.23001410067081451,
      "num_tokens": 31887292.0,
      "step": 13910
    },
    {
      "entropy": 5.294613647460937,
      "epoch": 1.3366954851104706,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004828000058891248,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22324256151914595,
      "num_tokens": 31899018.0,
      "step": 13915
    },
    {
      "entropy": 5.243601703643799,
      "epoch": 1.3371757925072045,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004827868569475706,
      "loss": 4.8325,
      "mean_token_accuracy": 0.23573760837316513,
      "num_tokens": 31910402.0,
      "step": 13920
    },
    {
      "entropy": 5.321239757537842,
      "epoch": 1.3376560999039384,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048277370318182243,
      "loss": 4.9698,
      "mean_token_accuracy": 0.21745618879795076,
      "num_tokens": 31921792.0,
      "step": 13925
    },
    {
      "entropy": 5.313501930236816,
      "epoch": 1.3381364073006723,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048276054459218596,
      "loss": 4.9622,
      "mean_token_accuracy": 0.22099509984254836,
      "num_tokens": 31932786.0,
      "step": 13930
    },
    {
      "entropy": 5.294550848007202,
      "epoch": 1.3386167146974064,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048274738117896643,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22700723260641098,
      "num_tokens": 31944286.0,
      "step": 13935
    },
    {
      "entropy": 5.264119243621826,
      "epoch": 1.3390970220941403,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048273421294246966,
      "loss": 4.9714,
      "mean_token_accuracy": 0.2197137862443924,
      "num_tokens": 31956304.0,
      "step": 13940
    },
    {
      "entropy": 5.243881034851074,
      "epoch": 1.3395773294908742,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048272103988300134,
      "loss": 4.8894,
      "mean_token_accuracy": 0.2287077859044075,
      "num_tokens": 31968978.0,
      "step": 13945
    },
    {
      "entropy": 5.301949167251587,
      "epoch": 1.340057636887608,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004827078620008672,
      "loss": 4.8783,
      "mean_token_accuracy": 0.22400756776332856,
      "num_tokens": 31980133.0,
      "step": 13950
    },
    {
      "entropy": 5.255629539489746,
      "epoch": 1.340537944284342,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048269467929637337,
      "loss": 4.8521,
      "mean_token_accuracy": 0.23400880098342897,
      "num_tokens": 31990331.0,
      "step": 13955
    },
    {
      "entropy": 5.3003133773803714,
      "epoch": 1.341018251681076,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048268149176982576,
      "loss": 4.8918,
      "mean_token_accuracy": 0.22627731710672377,
      "num_tokens": 32001956.0,
      "step": 13960
    },
    {
      "entropy": 5.319070434570312,
      "epoch": 1.34149855907781,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048266829942153055,
      "loss": 4.9421,
      "mean_token_accuracy": 0.2313989967107773,
      "num_tokens": 32013577.0,
      "step": 13965
    },
    {
      "entropy": 5.408236694335938,
      "epoch": 1.3419788664745438,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048265510225179413,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21986250281333924,
      "num_tokens": 32025751.0,
      "step": 13970
    },
    {
      "entropy": 5.283168029785156,
      "epoch": 1.3424591738712777,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004826419002609229,
      "loss": 4.9503,
      "mean_token_accuracy": 0.22096094936132432,
      "num_tokens": 32036592.0,
      "step": 13975
    },
    {
      "entropy": 5.355064344406128,
      "epoch": 1.3429394812680115,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048262869344922326,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22551192492246627,
      "num_tokens": 32048122.0,
      "step": 13980
    },
    {
      "entropy": 5.286335706710815,
      "epoch": 1.3434197886647454,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048261548181700186,
      "loss": 4.892,
      "mean_token_accuracy": 0.22413809448480607,
      "num_tokens": 32058673.0,
      "step": 13985
    },
    {
      "entropy": 5.191519165039063,
      "epoch": 1.3439000960614793,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004826022653645655,
      "loss": 4.8205,
      "mean_token_accuracy": 0.22988158464431763,
      "num_tokens": 32070219.0,
      "step": 13990
    },
    {
      "entropy": 5.317950010299683,
      "epoch": 1.3443804034582132,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000482589044092221,
      "loss": 4.9356,
      "mean_token_accuracy": 0.22349049299955367,
      "num_tokens": 32081883.0,
      "step": 13995
    },
    {
      "entropy": 5.373845958709717,
      "epoch": 1.344860710854947,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048257581800027527,
      "loss": 4.9706,
      "mean_token_accuracy": 0.21842403560876847,
      "num_tokens": 32094107.0,
      "step": 14000
    },
    {
      "entropy": 5.287773323059082,
      "epoch": 1.345341018251681,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004825625870890354,
      "loss": 4.8393,
      "mean_token_accuracy": 0.2288374736905098,
      "num_tokens": 32105242.0,
      "step": 14005
    },
    {
      "entropy": 5.284841680526734,
      "epoch": 1.345821325648415,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004825493513588086,
      "loss": 4.8405,
      "mean_token_accuracy": 0.23294240981340408,
      "num_tokens": 32117326.0,
      "step": 14010
    },
    {
      "entropy": 5.282085371017456,
      "epoch": 1.346301633045149,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048253611080990226,
      "loss": 4.9269,
      "mean_token_accuracy": 0.22332257330417632,
      "num_tokens": 32129002.0,
      "step": 14015
    },
    {
      "entropy": 5.2889251708984375,
      "epoch": 1.3467819404418828,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004825228654426236,
      "loss": 4.939,
      "mean_token_accuracy": 0.22172962576150895,
      "num_tokens": 32140380.0,
      "step": 14020
    },
    {
      "entropy": 5.244892883300781,
      "epoch": 1.3472622478386167,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004825096152572803,
      "loss": 4.8047,
      "mean_token_accuracy": 0.23440185338258743,
      "num_tokens": 32151806.0,
      "step": 14025
    },
    {
      "entropy": 5.302847576141358,
      "epoch": 1.3477425552353506,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048249636025417974,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2193058267235756,
      "num_tokens": 32163426.0,
      "step": 14030
    },
    {
      "entropy": 5.289321374893189,
      "epoch": 1.3482228626320845,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048248310043362997,
      "loss": 4.8879,
      "mean_token_accuracy": 0.23333536684513093,
      "num_tokens": 32174349.0,
      "step": 14035
    },
    {
      "entropy": 5.3538243770599365,
      "epoch": 1.3487031700288186,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004824698357959386,
      "loss": 4.9645,
      "mean_token_accuracy": 0.2250779390335083,
      "num_tokens": 32185382.0,
      "step": 14040
    },
    {
      "entropy": 5.303445100784302,
      "epoch": 1.3491834774255524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048245656634141385,
      "loss": 4.8569,
      "mean_token_accuracy": 0.23222119659185408,
      "num_tokens": 32195687.0,
      "step": 14045
    },
    {
      "entropy": 5.256520748138428,
      "epoch": 1.3496637848222863,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048244329207036354,
      "loss": 4.8206,
      "mean_token_accuracy": 0.23486315757036208,
      "num_tokens": 32205973.0,
      "step": 14050
    },
    {
      "entropy": 5.253281593322754,
      "epoch": 1.3501440922190202,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048243001298309604,
      "loss": 4.8926,
      "mean_token_accuracy": 0.22861236929893494,
      "num_tokens": 32217469.0,
      "step": 14055
    },
    {
      "entropy": 5.271201515197754,
      "epoch": 1.350624399615754,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048241672907991954,
      "loss": 4.9227,
      "mean_token_accuracy": 0.22756133526563643,
      "num_tokens": 32228257.0,
      "step": 14060
    },
    {
      "entropy": 5.2934332370758055,
      "epoch": 1.351104707012488,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004824034403611424,
      "loss": 4.884,
      "mean_token_accuracy": 0.23043331354856492,
      "num_tokens": 32239420.0,
      "step": 14065
    },
    {
      "entropy": 5.319902563095093,
      "epoch": 1.3515850144092219,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004823901468270733,
      "loss": 4.9146,
      "mean_token_accuracy": 0.22791532725095748,
      "num_tokens": 32250962.0,
      "step": 14070
    },
    {
      "entropy": 5.258719348907471,
      "epoch": 1.3520653218059557,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004823768484780209,
      "loss": 4.9068,
      "mean_token_accuracy": 0.23345289826393129,
      "num_tokens": 32262310.0,
      "step": 14075
    },
    {
      "entropy": 5.224226236343384,
      "epoch": 1.3525456292026896,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048236354531429375,
      "loss": 4.809,
      "mean_token_accuracy": 0.23438361436128616,
      "num_tokens": 32273373.0,
      "step": 14080
    },
    {
      "entropy": 5.3298135757446286,
      "epoch": 1.3530259365994235,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004823502373362009,
      "loss": 4.9288,
      "mean_token_accuracy": 0.22433389723300934,
      "num_tokens": 32285020.0,
      "step": 14085
    },
    {
      "entropy": 5.3010352611541744,
      "epoch": 1.3535062439961576,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004823369245440512,
      "loss": 4.9149,
      "mean_token_accuracy": 0.224274380505085,
      "num_tokens": 32296224.0,
      "step": 14090
    },
    {
      "entropy": 5.279384899139404,
      "epoch": 1.3539865513928915,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048232360693815387,
      "loss": 4.8959,
      "mean_token_accuracy": 0.23078217953443528,
      "num_tokens": 32306913.0,
      "step": 14095
    },
    {
      "entropy": 5.300556945800781,
      "epoch": 1.3544668587896254,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048231028451881786,
      "loss": 4.862,
      "mean_token_accuracy": 0.22505375742912292,
      "num_tokens": 32317981.0,
      "step": 14100
    },
    {
      "entropy": 5.18036003112793,
      "epoch": 1.3549471661863592,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004822969572863527,
      "loss": 4.7712,
      "mean_token_accuracy": 0.23677171021699905,
      "num_tokens": 32329656.0,
      "step": 14105
    },
    {
      "entropy": 5.291463470458984,
      "epoch": 1.3554274735830931,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048228362524106776,
      "loss": 4.9529,
      "mean_token_accuracy": 0.22679369747638703,
      "num_tokens": 32339460.0,
      "step": 14110
    },
    {
      "entropy": 5.24078402519226,
      "epoch": 1.3559077809798272,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048227028838327253,
      "loss": 4.8796,
      "mean_token_accuracy": 0.22564128637313843,
      "num_tokens": 32351237.0,
      "step": 14115
    },
    {
      "entropy": 5.2995072364807125,
      "epoch": 1.356388088376561,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048225694671327665,
      "loss": 4.814,
      "mean_token_accuracy": 0.22781677693128585,
      "num_tokens": 32362368.0,
      "step": 14120
    },
    {
      "entropy": 5.193451690673828,
      "epoch": 1.356868395773295,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004822436002313899,
      "loss": 4.8328,
      "mean_token_accuracy": 0.23790732622146607,
      "num_tokens": 32373738.0,
      "step": 14125
    },
    {
      "entropy": 5.254542684555053,
      "epoch": 1.3573487031700289,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004822302489379222,
      "loss": 4.9021,
      "mean_token_accuracy": 0.22736021429300307,
      "num_tokens": 32384274.0,
      "step": 14130
    },
    {
      "entropy": 5.219064664840698,
      "epoch": 1.3578290105667628,
      "grad_norm": 1.0,
      "learning_rate": 0.00048221689283318335,
      "loss": 4.8074,
      "mean_token_accuracy": 0.23711990416049958,
      "num_tokens": 32395692.0,
      "step": 14135
    },
    {
      "entropy": 5.225104904174804,
      "epoch": 1.3583093179634966,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004822035319174837,
      "loss": 4.8827,
      "mean_token_accuracy": 0.22602057009935378,
      "num_tokens": 32407367.0,
      "step": 14140
    },
    {
      "entropy": 5.285318183898926,
      "epoch": 1.3587896253602305,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004821901661911332,
      "loss": 4.9015,
      "mean_token_accuracy": 0.22378645688295365,
      "num_tokens": 32418753.0,
      "step": 14145
    },
    {
      "entropy": 5.3254087448120115,
      "epoch": 1.3592699327569644,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004821767956544423,
      "loss": 4.8945,
      "mean_token_accuracy": 0.22688398361206055,
      "num_tokens": 32429499.0,
      "step": 14150
    },
    {
      "entropy": 5.242872810363769,
      "epoch": 1.3597502401536983,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004821634203077214,
      "loss": 4.8529,
      "mean_token_accuracy": 0.22737785577774047,
      "num_tokens": 32441246.0,
      "step": 14155
    },
    {
      "entropy": 5.280128622055054,
      "epoch": 1.3602305475504322,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000482150040151281,
      "loss": 4.9276,
      "mean_token_accuracy": 0.22176198959350585,
      "num_tokens": 32452953.0,
      "step": 14160
    },
    {
      "entropy": 5.325202465057373,
      "epoch": 1.3607108549471663,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004821366551854318,
      "loss": 4.894,
      "mean_token_accuracy": 0.22397156357765197,
      "num_tokens": 32464030.0,
      "step": 14165
    },
    {
      "entropy": 5.2443320751190186,
      "epoch": 1.3611911623439001,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004821232654104845,
      "loss": 4.7077,
      "mean_token_accuracy": 0.23750038743019103,
      "num_tokens": 32475091.0,
      "step": 14170
    },
    {
      "entropy": 5.253853416442871,
      "epoch": 1.361671469740634,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048210987082675005,
      "loss": 4.9497,
      "mean_token_accuracy": 0.22417286932468414,
      "num_tokens": 32486047.0,
      "step": 14175
    },
    {
      "entropy": 5.247628402709961,
      "epoch": 1.362151777137368,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048209647143453946,
      "loss": 4.8207,
      "mean_token_accuracy": 0.23229122906923294,
      "num_tokens": 32497141.0,
      "step": 14180
    },
    {
      "entropy": 5.276901483535767,
      "epoch": 1.3626320845341018,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048208306723416356,
      "loss": 4.8658,
      "mean_token_accuracy": 0.23002825528383256,
      "num_tokens": 32509282.0,
      "step": 14185
    },
    {
      "entropy": 5.318138790130615,
      "epoch": 1.3631123919308357,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004820696582259339,
      "loss": 4.9504,
      "mean_token_accuracy": 0.22381894141435624,
      "num_tokens": 32521383.0,
      "step": 14190
    },
    {
      "entropy": 5.24918942451477,
      "epoch": 1.3635926993275698,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004820562444101616,
      "loss": 4.8414,
      "mean_token_accuracy": 0.2263380154967308,
      "num_tokens": 32533168.0,
      "step": 14195
    },
    {
      "entropy": 5.238666963577271,
      "epoch": 1.3640730067243036,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004820428257871581,
      "loss": 4.8786,
      "mean_token_accuracy": 0.22606067955493928,
      "num_tokens": 32545333.0,
      "step": 14200
    },
    {
      "entropy": 5.237027597427368,
      "epoch": 1.3645533141210375,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004820294023572351,
      "loss": 4.8377,
      "mean_token_accuracy": 0.23274746239185334,
      "num_tokens": 32556665.0,
      "step": 14205
    },
    {
      "entropy": 5.345625638961792,
      "epoch": 1.3650336215177714,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000482015974120704,
      "loss": 5.013,
      "mean_token_accuracy": 0.21710334420204164,
      "num_tokens": 32568318.0,
      "step": 14210
    },
    {
      "entropy": 5.27556791305542,
      "epoch": 1.3655139289145053,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048200254107787677,
      "loss": 4.7985,
      "mean_token_accuracy": 0.23009126782417297,
      "num_tokens": 32580010.0,
      "step": 14215
    },
    {
      "entropy": 5.307311773300171,
      "epoch": 1.3659942363112392,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048198910322906516,
      "loss": 4.9807,
      "mean_token_accuracy": 0.21753141433000564,
      "num_tokens": 32592075.0,
      "step": 14220
    },
    {
      "entropy": 5.385614538192749,
      "epoch": 1.366474543707973,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048197566057458125,
      "loss": 4.9965,
      "mean_token_accuracy": 0.22420198023319243,
      "num_tokens": 32604548.0,
      "step": 14225
    },
    {
      "entropy": 5.343211221694946,
      "epoch": 1.366954851104707,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000481962213114737,
      "loss": 4.9089,
      "mean_token_accuracy": 0.22370946705341338,
      "num_tokens": 32616036.0,
      "step": 14230
    },
    {
      "entropy": 5.220848798751831,
      "epoch": 1.3674351585014408,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004819487608498448,
      "loss": 4.8715,
      "mean_token_accuracy": 0.2226592019200325,
      "num_tokens": 32628157.0,
      "step": 14235
    },
    {
      "entropy": 5.309108686447144,
      "epoch": 1.3679154658981747,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048193530378021687,
      "loss": 4.9468,
      "mean_token_accuracy": 0.2259247660636902,
      "num_tokens": 32640571.0,
      "step": 14240
    },
    {
      "entropy": 5.295064687728882,
      "epoch": 1.3683957732949088,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048192184190616567,
      "loss": 4.9198,
      "mean_token_accuracy": 0.2238365665078163,
      "num_tokens": 32652005.0,
      "step": 14245
    },
    {
      "entropy": 5.281712579727173,
      "epoch": 1.3688760806916427,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004819083752280037,
      "loss": 4.8493,
      "mean_token_accuracy": 0.2311792716383934,
      "num_tokens": 32663655.0,
      "step": 14250
    },
    {
      "entropy": 5.2173622131347654,
      "epoch": 1.3693563880883766,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048189490374604373,
      "loss": 4.933,
      "mean_token_accuracy": 0.21644871681928635,
      "num_tokens": 32675419.0,
      "step": 14255
    },
    {
      "entropy": 5.255681848526001,
      "epoch": 1.3698366954851104,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004818814274605983,
      "loss": 4.8365,
      "mean_token_accuracy": 0.22816795110702515,
      "num_tokens": 32686680.0,
      "step": 14260
    },
    {
      "entropy": 5.239877223968506,
      "epoch": 1.3703170028818443,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004818679463719805,
      "loss": 4.7679,
      "mean_token_accuracy": 0.23506428450345992,
      "num_tokens": 32697321.0,
      "step": 14265
    },
    {
      "entropy": 5.196896505355835,
      "epoch": 1.3707973102785782,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004818544604805033,
      "loss": 4.8805,
      "mean_token_accuracy": 0.2230699986219406,
      "num_tokens": 32708885.0,
      "step": 14270
    },
    {
      "entropy": 5.30233588218689,
      "epoch": 1.3712776176753123,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004818409697864798,
      "loss": 4.8526,
      "mean_token_accuracy": 0.23407933712005616,
      "num_tokens": 32720517.0,
      "step": 14275
    },
    {
      "entropy": 5.253089380264282,
      "epoch": 1.3717579250720462,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048182747429022303,
      "loss": 4.8694,
      "mean_token_accuracy": 0.22458502650260925,
      "num_tokens": 32731072.0,
      "step": 14280
    },
    {
      "entropy": 5.352102899551392,
      "epoch": 1.37223823246878,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004818139739920465,
      "loss": 4.944,
      "mean_token_accuracy": 0.22814363092184067,
      "num_tokens": 32742117.0,
      "step": 14285
    },
    {
      "entropy": 5.28577184677124,
      "epoch": 1.372718539865514,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004818004688922637,
      "loss": 4.906,
      "mean_token_accuracy": 0.2242374002933502,
      "num_tokens": 32753754.0,
      "step": 14290
    },
    {
      "entropy": 5.218409395217895,
      "epoch": 1.3731988472622478,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000481786958991188,
      "loss": 4.857,
      "mean_token_accuracy": 0.22972660958766938,
      "num_tokens": 32766511.0,
      "step": 14295
    },
    {
      "entropy": 5.379793262481689,
      "epoch": 1.3736791546589817,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048177344428913316,
      "loss": 4.928,
      "mean_token_accuracy": 0.22117694616317748,
      "num_tokens": 32777715.0,
      "step": 14300
    },
    {
      "entropy": 5.298245525360107,
      "epoch": 1.3741594620557156,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048175992478641293,
      "loss": 4.8707,
      "mean_token_accuracy": 0.2259179711341858,
      "num_tokens": 32789132.0,
      "step": 14305
    },
    {
      "entropy": 5.3582195281982425,
      "epoch": 1.3746397694524495,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004817464004833412,
      "loss": 4.9396,
      "mean_token_accuracy": 0.22886692136526107,
      "num_tokens": 32800439.0,
      "step": 14310
    },
    {
      "entropy": 5.274745798110962,
      "epoch": 1.3751200768491834,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048173287138023204,
      "loss": 4.8436,
      "mean_token_accuracy": 0.22915974855422974,
      "num_tokens": 32813605.0,
      "step": 14315
    },
    {
      "entropy": 5.356570720672607,
      "epoch": 1.3756003842459175,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004817193374773993,
      "loss": 4.9518,
      "mean_token_accuracy": 0.2287974089384079,
      "num_tokens": 32824225.0,
      "step": 14320
    },
    {
      "entropy": 5.349268913269043,
      "epoch": 1.3760806916426513,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048170579877515753,
      "loss": 5.0028,
      "mean_token_accuracy": 0.21382326632738113,
      "num_tokens": 32836917.0,
      "step": 14325
    },
    {
      "entropy": 5.262779617309571,
      "epoch": 1.3765609990393852,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004816922552738209,
      "loss": 4.8372,
      "mean_token_accuracy": 0.22766863703727722,
      "num_tokens": 32847972.0,
      "step": 14330
    },
    {
      "entropy": 5.169267034530639,
      "epoch": 1.377041306436119,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048167870697370373,
      "loss": 4.8184,
      "mean_token_accuracy": 0.23471040427684783,
      "num_tokens": 32858922.0,
      "step": 14335
    },
    {
      "entropy": 5.264252328872681,
      "epoch": 1.377521613832853,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004816651538751207,
      "loss": 4.7912,
      "mean_token_accuracy": 0.23346583247184755,
      "num_tokens": 32869788.0,
      "step": 14340
    },
    {
      "entropy": 5.420011949539185,
      "epoch": 1.3780019212295869,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048165159597838664,
      "loss": 5.0014,
      "mean_token_accuracy": 0.2160484179854393,
      "num_tokens": 32881678.0,
      "step": 14345
    },
    {
      "entropy": 5.314347648620606,
      "epoch": 1.378482228626321,
      "grad_norm": 0.953125,
      "learning_rate": 0.000481638033283816,
      "loss": 4.9066,
      "mean_token_accuracy": 0.22295429259538652,
      "num_tokens": 32894183.0,
      "step": 14350
    },
    {
      "entropy": 5.242513513565063,
      "epoch": 1.3789625360230549,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048162446579172387,
      "loss": 4.8492,
      "mean_token_accuracy": 0.23041439801454544,
      "num_tokens": 32906001.0,
      "step": 14355
    },
    {
      "entropy": 5.287830352783203,
      "epoch": 1.3794428434197887,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004816108935024252,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22436414510011674,
      "num_tokens": 32917236.0,
      "step": 14360
    },
    {
      "entropy": 5.283598899841309,
      "epoch": 1.3799231508165226,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048159731641623507,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22699737995862962,
      "num_tokens": 32929710.0,
      "step": 14365
    },
    {
      "entropy": 5.26976375579834,
      "epoch": 1.3804034582132565,
      "grad_norm": 1.125,
      "learning_rate": 0.0004815837345334687,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2283136159181595,
      "num_tokens": 32941565.0,
      "step": 14370
    },
    {
      "entropy": 5.368187665939331,
      "epoch": 1.3808837656099904,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004815701478544415,
      "loss": 5.0029,
      "mean_token_accuracy": 0.21981867402791977,
      "num_tokens": 32952730.0,
      "step": 14375
    },
    {
      "entropy": 5.261301708221436,
      "epoch": 1.3813640730067243,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048155655637946876,
      "loss": 4.836,
      "mean_token_accuracy": 0.23050251454114914,
      "num_tokens": 32963589.0,
      "step": 14380
    },
    {
      "entropy": 5.256851148605347,
      "epoch": 1.3818443804034581,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004815429601088662,
      "loss": 4.802,
      "mean_token_accuracy": 0.23001941442489623,
      "num_tokens": 32975008.0,
      "step": 14385
    },
    {
      "entropy": 5.302001190185547,
      "epoch": 1.382324687800192,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004815293590429494,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2169436112046242,
      "num_tokens": 32986497.0,
      "step": 14390
    },
    {
      "entropy": 5.217916822433471,
      "epoch": 1.382804995196926,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048151575318203417,
      "loss": 4.7355,
      "mean_token_accuracy": 0.23918542712926866,
      "num_tokens": 32998298.0,
      "step": 14395
    },
    {
      "entropy": 5.372474527359008,
      "epoch": 1.38328530259366,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048150214252643637,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2207875818014145,
      "num_tokens": 33010943.0,
      "step": 14400
    },
    {
      "entropy": 5.2929400444030765,
      "epoch": 1.3837656099903939,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000481488527076472,
      "loss": 4.8974,
      "mean_token_accuracy": 0.2264263466000557,
      "num_tokens": 33022139.0,
      "step": 14405
    },
    {
      "entropy": 5.327686309814453,
      "epoch": 1.3842459173871278,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004814749068324571,
      "loss": 4.9806,
      "mean_token_accuracy": 0.22565121352672576,
      "num_tokens": 33034653.0,
      "step": 14410
    },
    {
      "entropy": 5.198713493347168,
      "epoch": 1.3847262247838616,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048146128179470804,
      "loss": 4.6695,
      "mean_token_accuracy": 0.246149642765522,
      "num_tokens": 33044339.0,
      "step": 14415
    },
    {
      "entropy": 5.2544506072998045,
      "epoch": 1.3852065321805955,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004814476519635411,
      "loss": 4.8859,
      "mean_token_accuracy": 0.230136638879776,
      "num_tokens": 33055418.0,
      "step": 14420
    },
    {
      "entropy": 5.272707271575928,
      "epoch": 1.3856868395773294,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048143401733927274,
      "loss": 4.9725,
      "mean_token_accuracy": 0.22618777602910994,
      "num_tokens": 33067239.0,
      "step": 14425
    },
    {
      "entropy": 5.2397225856781,
      "epoch": 1.3861671469740635,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048142037792221943,
      "loss": 4.765,
      "mean_token_accuracy": 0.23663000464439393,
      "num_tokens": 33079101.0,
      "step": 14430
    },
    {
      "entropy": 5.306547355651856,
      "epoch": 1.3866474543707974,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004814067337126978,
      "loss": 4.8703,
      "mean_token_accuracy": 0.23314654380083083,
      "num_tokens": 33090265.0,
      "step": 14435
    },
    {
      "entropy": 5.264362144470215,
      "epoch": 1.3871277617675313,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004813930847110248,
      "loss": 4.825,
      "mean_token_accuracy": 0.23466628193855285,
      "num_tokens": 33101724.0,
      "step": 14440
    },
    {
      "entropy": 5.221672010421753,
      "epoch": 1.3876080691642652,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004813794309175173,
      "loss": 4.8041,
      "mean_token_accuracy": 0.23698410540819168,
      "num_tokens": 33113111.0,
      "step": 14445
    },
    {
      "entropy": 5.335699939727784,
      "epoch": 1.388088376560999,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048136577233249205,
      "loss": 4.9693,
      "mean_token_accuracy": 0.21737074702978135,
      "num_tokens": 33123925.0,
      "step": 14450
    },
    {
      "entropy": 5.303652811050415,
      "epoch": 1.388568683957733,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004813521089562666,
      "loss": 4.9184,
      "mean_token_accuracy": 0.22257649004459382,
      "num_tokens": 33137400.0,
      "step": 14455
    },
    {
      "entropy": 5.182322263717651,
      "epoch": 1.3890489913544668,
      "grad_norm": 0.875,
      "learning_rate": 0.0004813384407891577,
      "loss": 4.7692,
      "mean_token_accuracy": 0.23685832619667052,
      "num_tokens": 33149326.0,
      "step": 14460
    },
    {
      "entropy": 5.346370267868042,
      "epoch": 1.3895292987512007,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000481324767831483,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21195703893899917,
      "num_tokens": 33159904.0,
      "step": 14465
    },
    {
      "entropy": 5.237347221374511,
      "epoch": 1.3900096061479346,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004813110900835598,
      "loss": 4.8518,
      "mean_token_accuracy": 0.23176784813404083,
      "num_tokens": 33171701.0,
      "step": 14470
    },
    {
      "entropy": 5.284486532211304,
      "epoch": 1.3904899135446687,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004812974075457058,
      "loss": 4.7645,
      "mean_token_accuracy": 0.23253712058067322,
      "num_tokens": 33181598.0,
      "step": 14475
    },
    {
      "entropy": 5.249352931976318,
      "epoch": 1.3909702209414025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048128372021823845,
      "loss": 4.7922,
      "mean_token_accuracy": 0.23322942554950715,
      "num_tokens": 33192674.0,
      "step": 14480
    },
    {
      "entropy": 5.20747127532959,
      "epoch": 1.3914505283381364,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048127002810147574,
      "loss": 4.8573,
      "mean_token_accuracy": 0.23061423301696776,
      "num_tokens": 33203356.0,
      "step": 14485
    },
    {
      "entropy": 5.274855327606201,
      "epoch": 1.3919308357348703,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004812563311957355,
      "loss": 4.8855,
      "mean_token_accuracy": 0.22962510734796523,
      "num_tokens": 33215471.0,
      "step": 14490
    },
    {
      "entropy": 5.272241067886353,
      "epoch": 1.3924111431316042,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004812426295013356,
      "loss": 4.8678,
      "mean_token_accuracy": 0.22269374430179595,
      "num_tokens": 33227656.0,
      "step": 14495
    },
    {
      "entropy": 5.319069910049438,
      "epoch": 1.392891450528338,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048122892301859433,
      "loss": 4.9819,
      "mean_token_accuracy": 0.22536935955286025,
      "num_tokens": 33239752.0,
      "step": 14500
    },
    {
      "entropy": 5.322785568237305,
      "epoch": 1.3933717579250722,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048121521174782983,
      "loss": 4.9093,
      "mean_token_accuracy": 0.2229253813624382,
      "num_tokens": 33251352.0,
      "step": 14505
    },
    {
      "entropy": 5.300270128250122,
      "epoch": 1.393852065321806,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048120149568936044,
      "loss": 4.9029,
      "mean_token_accuracy": 0.22683097571134567,
      "num_tokens": 33262276.0,
      "step": 14510
    },
    {
      "entropy": 5.334196281433106,
      "epoch": 1.39433237271854,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004811877748435046,
      "loss": 4.9457,
      "mean_token_accuracy": 0.22120414972305297,
      "num_tokens": 33273615.0,
      "step": 14515
    },
    {
      "entropy": 5.281824541091919,
      "epoch": 1.3948126801152738,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004811740492105809,
      "loss": 4.862,
      "mean_token_accuracy": 0.23330594301223756,
      "num_tokens": 33283990.0,
      "step": 14520
    },
    {
      "entropy": 5.20621428489685,
      "epoch": 1.3952929875120077,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000481160318790908,
      "loss": 4.8725,
      "mean_token_accuracy": 0.23369529992341995,
      "num_tokens": 33295472.0,
      "step": 14525
    },
    {
      "entropy": 5.30248761177063,
      "epoch": 1.3957732949087416,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048114658358480467,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22270412743091583,
      "num_tokens": 33306742.0,
      "step": 14530
    },
    {
      "entropy": 5.3407790184021,
      "epoch": 1.3962536023054755,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048113284359258977,
      "loss": 4.9133,
      "mean_token_accuracy": 0.22399671375751495,
      "num_tokens": 33317737.0,
      "step": 14535
    },
    {
      "entropy": 5.305859184265136,
      "epoch": 1.3967339097022093,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048111909881458234,
      "loss": 4.9344,
      "mean_token_accuracy": 0.22172762602567672,
      "num_tokens": 33329673.0,
      "step": 14540
    },
    {
      "entropy": 5.167238664627075,
      "epoch": 1.3972142170989432,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048110534925110146,
      "loss": 4.7377,
      "mean_token_accuracy": 0.23782808035612107,
      "num_tokens": 33342047.0,
      "step": 14545
    },
    {
      "entropy": 5.22300534248352,
      "epoch": 1.397694524495677,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004810915949024664,
      "loss": 4.8428,
      "mean_token_accuracy": 0.2351212441921234,
      "num_tokens": 33353287.0,
      "step": 14550
    },
    {
      "entropy": 5.314918947219849,
      "epoch": 1.3981748318924112,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004810778357689965,
      "loss": 4.9245,
      "mean_token_accuracy": 0.23019034117460252,
      "num_tokens": 33365465.0,
      "step": 14555
    },
    {
      "entropy": 5.261622619628906,
      "epoch": 1.398655139289145,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048106407185101116,
      "loss": 4.8614,
      "mean_token_accuracy": 0.22797708064317704,
      "num_tokens": 33376680.0,
      "step": 14560
    },
    {
      "entropy": 5.229944896697998,
      "epoch": 1.399135446685879,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048105030314883,
      "loss": 4.8661,
      "mean_token_accuracy": 0.22796406596899033,
      "num_tokens": 33387995.0,
      "step": 14565
    },
    {
      "entropy": 5.263455200195312,
      "epoch": 1.3996157540826129,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004810365296627725,
      "loss": 4.8253,
      "mean_token_accuracy": 0.23222047835588455,
      "num_tokens": 33400455.0,
      "step": 14570
    },
    {
      "entropy": 5.323958873748779,
      "epoch": 1.4000960614793467,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004810227513931587,
      "loss": 4.8935,
      "mean_token_accuracy": 0.22289664596319197,
      "num_tokens": 33413264.0,
      "step": 14575
    },
    {
      "entropy": 5.240088033676147,
      "epoch": 1.4005763688760806,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004810089683403084,
      "loss": 4.774,
      "mean_token_accuracy": 0.23885849863290787,
      "num_tokens": 33423516.0,
      "step": 14580
    },
    {
      "entropy": 5.26455979347229,
      "epoch": 1.4010566762728147,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004809951805045415,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2193303033709526,
      "num_tokens": 33434952.0,
      "step": 14585
    },
    {
      "entropy": 5.26010479927063,
      "epoch": 1.4015369836695486,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048098138788617815,
      "loss": 4.8724,
      "mean_token_accuracy": 0.22798535823822022,
      "num_tokens": 33447025.0,
      "step": 14590
    },
    {
      "entropy": 5.299524211883545,
      "epoch": 1.4020172910662825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004809675904855387,
      "loss": 4.8522,
      "mean_token_accuracy": 0.22846070677042007,
      "num_tokens": 33459990.0,
      "step": 14595
    },
    {
      "entropy": 5.262665843963623,
      "epoch": 1.4024975984630164,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048095378830294343,
      "loss": 4.8009,
      "mean_token_accuracy": 0.23399470001459122,
      "num_tokens": 33471334.0,
      "step": 14600
    },
    {
      "entropy": 5.294054698944092,
      "epoch": 1.4029779058597502,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048093998133871276,
      "loss": 4.9169,
      "mean_token_accuracy": 0.22734078615903855,
      "num_tokens": 33483744.0,
      "step": 14605
    },
    {
      "entropy": 5.277678489685059,
      "epoch": 1.4034582132564841,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004809261695931671,
      "loss": 4.9535,
      "mean_token_accuracy": 0.2225761115550995,
      "num_tokens": 33496532.0,
      "step": 14610
    },
    {
      "entropy": 5.218972158432007,
      "epoch": 1.403938520653218,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004809123530666273,
      "loss": 4.8704,
      "mean_token_accuracy": 0.23326461166143417,
      "num_tokens": 33507545.0,
      "step": 14615
    },
    {
      "entropy": 5.30824613571167,
      "epoch": 1.4044188280499519,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004808985317594142,
      "loss": 4.8605,
      "mean_token_accuracy": 0.23201444894075393,
      "num_tokens": 33519232.0,
      "step": 14620
    },
    {
      "entropy": 5.26504168510437,
      "epoch": 1.4048991354466858,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048088470567184854,
      "loss": 4.843,
      "mean_token_accuracy": 0.22826878279447554,
      "num_tokens": 33531088.0,
      "step": 14625
    },
    {
      "entropy": 5.302885246276856,
      "epoch": 1.4053794428434199,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048087087480425133,
      "loss": 4.909,
      "mean_token_accuracy": 0.22353773415088654,
      "num_tokens": 33543991.0,
      "step": 14630
    },
    {
      "entropy": 5.353040361404419,
      "epoch": 1.4058597502401537,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004808570391569437,
      "loss": 4.9588,
      "mean_token_accuracy": 0.21770821511745453,
      "num_tokens": 33555303.0,
      "step": 14635
    },
    {
      "entropy": 5.324752473831177,
      "epoch": 1.4063400576368876,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048084319873024694,
      "loss": 4.9331,
      "mean_token_accuracy": 0.22396344393491746,
      "num_tokens": 33565587.0,
      "step": 14640
    },
    {
      "entropy": 5.266775321960449,
      "epoch": 1.4068203650336215,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004808293535244823,
      "loss": 4.8006,
      "mean_token_accuracy": 0.22821844220161439,
      "num_tokens": 33576874.0,
      "step": 14645
    },
    {
      "entropy": 5.243024587631226,
      "epoch": 1.4073006724303554,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004808155035399712,
      "loss": 4.822,
      "mean_token_accuracy": 0.23688398897647858,
      "num_tokens": 33587703.0,
      "step": 14650
    },
    {
      "entropy": 5.20507025718689,
      "epoch": 1.4077809798270893,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004808016487770354,
      "loss": 4.7835,
      "mean_token_accuracy": 0.23999144732952118,
      "num_tokens": 33598487.0,
      "step": 14655
    },
    {
      "entropy": 5.252055406570435,
      "epoch": 1.4082612872238234,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048078778923599637,
      "loss": 4.9202,
      "mean_token_accuracy": 0.22637830674648285,
      "num_tokens": 33610838.0,
      "step": 14660
    },
    {
      "entropy": 5.2665668487548825,
      "epoch": 1.4087415946205573,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048077392491717593,
      "loss": 4.8674,
      "mean_token_accuracy": 0.2270650163292885,
      "num_tokens": 33622726.0,
      "step": 14665
    },
    {
      "entropy": 5.2252280712127686,
      "epoch": 1.4092219020172911,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048076005582089597,
      "loss": 4.8362,
      "mean_token_accuracy": 0.23020372837781905,
      "num_tokens": 33635922.0,
      "step": 14670
    },
    {
      "entropy": 5.37718391418457,
      "epoch": 1.409702209414025,
      "grad_norm": 1.0,
      "learning_rate": 0.00048074618194747845,
      "loss": 4.9746,
      "mean_token_accuracy": 0.22299656122922898,
      "num_tokens": 33648486.0,
      "step": 14675
    },
    {
      "entropy": 5.308882093429565,
      "epoch": 1.410182516810759,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004807323032972456,
      "loss": 4.8846,
      "mean_token_accuracy": 0.22957604229450226,
      "num_tokens": 33659812.0,
      "step": 14680
    },
    {
      "entropy": 5.292644453048706,
      "epoch": 1.4106628242074928,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004807184198705196,
      "loss": 4.8663,
      "mean_token_accuracy": 0.23246962279081346,
      "num_tokens": 33671878.0,
      "step": 14685
    },
    {
      "entropy": 5.1807191371917725,
      "epoch": 1.4111431316042267,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004807045316676226,
      "loss": 4.7781,
      "mean_token_accuracy": 0.2314123660326004,
      "num_tokens": 33683759.0,
      "step": 14690
    },
    {
      "entropy": 5.341965198516846,
      "epoch": 1.4116234390009605,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004806906386888773,
      "loss": 4.9801,
      "mean_token_accuracy": 0.21838423907756804,
      "num_tokens": 33694085.0,
      "step": 14695
    },
    {
      "entropy": 5.272063970565796,
      "epoch": 1.4121037463976944,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048067674093460607,
      "loss": 4.8198,
      "mean_token_accuracy": 0.2293445274233818,
      "num_tokens": 33705810.0,
      "step": 14700
    },
    {
      "entropy": 5.305046463012696,
      "epoch": 1.4125840537944283,
      "grad_norm": 1.0,
      "learning_rate": 0.00048066283840513175,
      "loss": 4.9128,
      "mean_token_accuracy": 0.2288290038704872,
      "num_tokens": 33716798.0,
      "step": 14705
    },
    {
      "entropy": 5.246756219863892,
      "epoch": 1.4130643611911624,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004806489311007769,
      "loss": 4.871,
      "mean_token_accuracy": 0.23461257070302963,
      "num_tokens": 33728515.0,
      "step": 14710
    },
    {
      "entropy": 5.27306809425354,
      "epoch": 1.4135446685878963,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048063501902186463,
      "loss": 4.881,
      "mean_token_accuracy": 0.22897555381059648,
      "num_tokens": 33740684.0,
      "step": 14715
    },
    {
      "entropy": 5.253626775741577,
      "epoch": 1.4140249759846302,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048062110216871775,
      "loss": 4.8594,
      "mean_token_accuracy": 0.23071405440568923,
      "num_tokens": 33753017.0,
      "step": 14720
    },
    {
      "entropy": 5.268177843093872,
      "epoch": 1.414505283381364,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048060718054165945,
      "loss": 4.8767,
      "mean_token_accuracy": 0.22858474105596543,
      "num_tokens": 33764393.0,
      "step": 14725
    },
    {
      "entropy": 5.330766105651856,
      "epoch": 1.414985590778098,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000480593254141013,
      "loss": 4.9976,
      "mean_token_accuracy": 0.21258441358804703,
      "num_tokens": 33774941.0,
      "step": 14730
    },
    {
      "entropy": 5.179855680465698,
      "epoch": 1.4154658981748318,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048057932296710165,
      "loss": 4.7517,
      "mean_token_accuracy": 0.2439322903752327,
      "num_tokens": 33786534.0,
      "step": 14735
    },
    {
      "entropy": 5.374244976043701,
      "epoch": 1.415946205571566,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004805653870202489,
      "loss": 4.9949,
      "mean_token_accuracy": 0.21618150770664216,
      "num_tokens": 33798339.0,
      "step": 14740
    },
    {
      "entropy": 5.385513782501221,
      "epoch": 1.4164265129682998,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048055144630077825,
      "loss": 5.0198,
      "mean_token_accuracy": 0.21582156866788865,
      "num_tokens": 33810368.0,
      "step": 14745
    },
    {
      "entropy": 5.3310973167419435,
      "epoch": 1.4169068203650337,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048053750080901336,
      "loss": 4.861,
      "mean_token_accuracy": 0.22740892171859742,
      "num_tokens": 33821111.0,
      "step": 14750
    },
    {
      "entropy": 5.2287187576293945,
      "epoch": 1.4173871277617676,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048052355054527794,
      "loss": 4.8562,
      "mean_token_accuracy": 0.22929963767528533,
      "num_tokens": 33833629.0,
      "step": 14755
    },
    {
      "entropy": 5.2800078868865965,
      "epoch": 1.4178674351585014,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048050959550989606,
      "loss": 4.9389,
      "mean_token_accuracy": 0.2239015594124794,
      "num_tokens": 33846531.0,
      "step": 14760
    },
    {
      "entropy": 5.228882646560669,
      "epoch": 1.4183477425552353,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004804956357031916,
      "loss": 4.8559,
      "mean_token_accuracy": 0.22688696533441544,
      "num_tokens": 33857251.0,
      "step": 14765
    },
    {
      "entropy": 5.263487148284912,
      "epoch": 1.4188280499519692,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048048167112548873,
      "loss": 4.8507,
      "mean_token_accuracy": 0.22788214087486267,
      "num_tokens": 33869314.0,
      "step": 14770
    },
    {
      "entropy": 5.354525184631347,
      "epoch": 1.419308357348703,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048046770177711157,
      "loss": 4.9658,
      "mean_token_accuracy": 0.22174745202064514,
      "num_tokens": 33880203.0,
      "step": 14775
    },
    {
      "entropy": 5.257967615127564,
      "epoch": 1.419788664745437,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004804537276583844,
      "loss": 4.7927,
      "mean_token_accuracy": 0.23879806250333785,
      "num_tokens": 33891379.0,
      "step": 14780
    },
    {
      "entropy": 5.232910585403443,
      "epoch": 1.420268972142171,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004804397487696319,
      "loss": 4.8304,
      "mean_token_accuracy": 0.23079285621643067,
      "num_tokens": 33902788.0,
      "step": 14785
    },
    {
      "entropy": 5.233991241455078,
      "epoch": 1.420749279538905,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004804257651111783,
      "loss": 4.8722,
      "mean_token_accuracy": 0.2327768623828888,
      "num_tokens": 33913609.0,
      "step": 14790
    },
    {
      "entropy": 5.262866830825805,
      "epoch": 1.4212295869356388,
      "grad_norm": 1.0,
      "learning_rate": 0.00048041177668334853,
      "loss": 4.786,
      "mean_token_accuracy": 0.23209329843521118,
      "num_tokens": 33924379.0,
      "step": 14795
    },
    {
      "entropy": 5.272022294998169,
      "epoch": 1.4217098943323727,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004803977834864672,
      "loss": 4.8978,
      "mean_token_accuracy": 0.22704977989196778,
      "num_tokens": 33936209.0,
      "step": 14800
    },
    {
      "entropy": 5.3395716667175295,
      "epoch": 1.4221902017291066,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048038378552085927,
      "loss": 4.865,
      "mean_token_accuracy": 0.21809807419776917,
      "num_tokens": 33947679.0,
      "step": 14805
    },
    {
      "entropy": 5.265112686157226,
      "epoch": 1.4226705091258405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048036978278684974,
      "loss": 4.8699,
      "mean_token_accuracy": 0.2272411420941353,
      "num_tokens": 33959474.0,
      "step": 14810
    },
    {
      "entropy": 5.373313856124878,
      "epoch": 1.4231508165225746,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004803557752847636,
      "loss": 5.0096,
      "mean_token_accuracy": 0.21955397427082063,
      "num_tokens": 33970831.0,
      "step": 14815
    },
    {
      "entropy": 5.389080667495728,
      "epoch": 1.4236311239193085,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048034176301492616,
      "loss": 4.9533,
      "mean_token_accuracy": 0.2223074406385422,
      "num_tokens": 33981765.0,
      "step": 14820
    },
    {
      "entropy": 5.298489618301391,
      "epoch": 1.4241114313160423,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004803277459776628,
      "loss": 4.8559,
      "mean_token_accuracy": 0.225976100564003,
      "num_tokens": 33992435.0,
      "step": 14825
    },
    {
      "entropy": 5.19735746383667,
      "epoch": 1.4245917387127762,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048031372417329875,
      "loss": 4.8291,
      "mean_token_accuracy": 0.2356481894850731,
      "num_tokens": 34004570.0,
      "step": 14830
    },
    {
      "entropy": 5.261145448684692,
      "epoch": 1.42507204610951,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004802996976021598,
      "loss": 4.8118,
      "mean_token_accuracy": 0.2338303357362747,
      "num_tokens": 34015494.0,
      "step": 14835
    },
    {
      "entropy": 5.333854484558105,
      "epoch": 1.425552353506244,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048028566626457145,
      "loss": 4.8981,
      "mean_token_accuracy": 0.22211015373468398,
      "num_tokens": 34026684.0,
      "step": 14840
    },
    {
      "entropy": 5.26174030303955,
      "epoch": 1.4260326609029779,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048027163016085947,
      "loss": 4.8631,
      "mean_token_accuracy": 0.22287061363458632,
      "num_tokens": 34038948.0,
      "step": 14845
    },
    {
      "entropy": 5.319403266906738,
      "epoch": 1.4265129682997117,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048025758929134976,
      "loss": 5.012,
      "mean_token_accuracy": 0.2161506935954094,
      "num_tokens": 34052216.0,
      "step": 14850
    },
    {
      "entropy": 5.311048364639282,
      "epoch": 1.4269932756964456,
      "grad_norm": 1.0,
      "learning_rate": 0.0004802435436563684,
      "loss": 4.8843,
      "mean_token_accuracy": 0.23151980191469193,
      "num_tokens": 34062602.0,
      "step": 14855
    },
    {
      "entropy": 5.314554023742676,
      "epoch": 1.4274735830931795,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048022949325624134,
      "loss": 4.8542,
      "mean_token_accuracy": 0.23099010288715363,
      "num_tokens": 34075049.0,
      "step": 14860
    },
    {
      "entropy": 5.303242588043213,
      "epoch": 1.4279538904899136,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048021543809129483,
      "loss": 4.9378,
      "mean_token_accuracy": 0.2208562210202217,
      "num_tokens": 34086719.0,
      "step": 14865
    },
    {
      "entropy": 5.31861138343811,
      "epoch": 1.4284341978866475,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004802013781618552,
      "loss": 4.8541,
      "mean_token_accuracy": 0.2285666972398758,
      "num_tokens": 34098439.0,
      "step": 14870
    },
    {
      "entropy": 5.347563457489014,
      "epoch": 1.4289145052833814,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048018731346824895,
      "loss": 4.8706,
      "mean_token_accuracy": 0.22942040264606475,
      "num_tokens": 34110711.0,
      "step": 14875
    },
    {
      "entropy": 5.202267074584961,
      "epoch": 1.4293948126801153,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004801732440108026,
      "loss": 4.8279,
      "mean_token_accuracy": 0.22897837907075883,
      "num_tokens": 34122191.0,
      "step": 14880
    },
    {
      "entropy": 5.179712295532227,
      "epoch": 1.4298751200768491,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004801591697898427,
      "loss": 4.7886,
      "mean_token_accuracy": 0.23265341967344283,
      "num_tokens": 34132838.0,
      "step": 14885
    },
    {
      "entropy": 5.323724937438965,
      "epoch": 1.430355427473583,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004801450908056961,
      "loss": 4.9095,
      "mean_token_accuracy": 0.2169827938079834,
      "num_tokens": 34143394.0,
      "step": 14890
    },
    {
      "entropy": 5.403874111175537,
      "epoch": 1.4308357348703171,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004801310070586896,
      "loss": 4.963,
      "mean_token_accuracy": 0.2208109751343727,
      "num_tokens": 34155934.0,
      "step": 14895
    },
    {
      "entropy": 5.131343936920166,
      "epoch": 1.431316042267051,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004801169185491503,
      "loss": 4.7504,
      "mean_token_accuracy": 0.23865431249141694,
      "num_tokens": 34167949.0,
      "step": 14900
    },
    {
      "entropy": 5.344604921340943,
      "epoch": 1.4317963496637849,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048010282527740516,
      "loss": 5.0214,
      "mean_token_accuracy": 0.21716126799583435,
      "num_tokens": 34179733.0,
      "step": 14905
    },
    {
      "entropy": 5.286039209365844,
      "epoch": 1.4322766570605188,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048008872724378146,
      "loss": 4.8016,
      "mean_token_accuracy": 0.229231296479702,
      "num_tokens": 34190513.0,
      "step": 14910
    },
    {
      "entropy": 5.235869312286377,
      "epoch": 1.4327569644572526,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004800746244486065,
      "loss": 4.8356,
      "mean_token_accuracy": 0.23347446471452712,
      "num_tokens": 34201562.0,
      "step": 14915
    },
    {
      "entropy": 5.1842913150787355,
      "epoch": 1.4332372718539865,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004800605168922077,
      "loss": 4.8089,
      "mean_token_accuracy": 0.2300594210624695,
      "num_tokens": 34212637.0,
      "step": 14920
    },
    {
      "entropy": 5.233330106735229,
      "epoch": 1.4337175792507204,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048004640457491267,
      "loss": 4.8484,
      "mean_token_accuracy": 0.22853269129991532,
      "num_tokens": 34225394.0,
      "step": 14925
    },
    {
      "entropy": 5.311647558212281,
      "epoch": 1.4341978866474543,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000480032287497049,
      "loss": 4.8408,
      "mean_token_accuracy": 0.2269302174448967,
      "num_tokens": 34236977.0,
      "step": 14930
    },
    {
      "entropy": 5.298158597946167,
      "epoch": 1.4346781940441882,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048001816565894427,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22857412844896316,
      "num_tokens": 34247486.0,
      "step": 14935
    },
    {
      "entropy": 5.259630823135376,
      "epoch": 1.435158501440922,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004800040390609267,
      "loss": 4.8282,
      "mean_token_accuracy": 0.23835201412439347,
      "num_tokens": 34259404.0,
      "step": 14940
    },
    {
      "entropy": 5.249775123596192,
      "epoch": 1.4356388088376562,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047998990770332396,
      "loss": 4.822,
      "mean_token_accuracy": 0.23836783319711685,
      "num_tokens": 34270388.0,
      "step": 14945
    },
    {
      "entropy": 5.345383358001709,
      "epoch": 1.43611911623439,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004799757715864643,
      "loss": 4.8958,
      "mean_token_accuracy": 0.22910062223672867,
      "num_tokens": 34281321.0,
      "step": 14950
    },
    {
      "entropy": 5.369006729125976,
      "epoch": 1.436599423631124,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004799616307106759,
      "loss": 4.9451,
      "mean_token_accuracy": 0.2159550219774246,
      "num_tokens": 34293177.0,
      "step": 14955
    },
    {
      "entropy": 5.316050910949707,
      "epoch": 1.4370797310278578,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000479947485076287,
      "loss": 4.9269,
      "mean_token_accuracy": 0.22394849509000778,
      "num_tokens": 34305175.0,
      "step": 14960
    },
    {
      "entropy": 5.285770130157471,
      "epoch": 1.4375600384245917,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047993333468362607,
      "loss": 4.9268,
      "mean_token_accuracy": 0.2192586436867714,
      "num_tokens": 34317969.0,
      "step": 14965
    },
    {
      "entropy": 5.393619585037231,
      "epoch": 1.4380403458213258,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047991917953302173,
      "loss": 4.946,
      "mean_token_accuracy": 0.22146107852458954,
      "num_tokens": 34329913.0,
      "step": 14970
    },
    {
      "entropy": 5.359150123596192,
      "epoch": 1.4385206532180597,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047990501962480236,
      "loss": 4.8876,
      "mean_token_accuracy": 0.22894523441791534,
      "num_tokens": 34341656.0,
      "step": 14975
    },
    {
      "entropy": 5.270420932769776,
      "epoch": 1.4390009606147935,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000479890854959297,
      "loss": 4.8559,
      "mean_token_accuracy": 0.22723519057035446,
      "num_tokens": 34351767.0,
      "step": 14980
    },
    {
      "entropy": 5.248279237747193,
      "epoch": 1.4394812680115274,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004798766855368344,
      "loss": 4.8533,
      "mean_token_accuracy": 0.22984237372875213,
      "num_tokens": 34363437.0,
      "step": 14985
    },
    {
      "entropy": 5.3803709030151365,
      "epoch": 1.4399615754082613,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047986251135774343,
      "loss": 4.9285,
      "mean_token_accuracy": 0.22940934151411058,
      "num_tokens": 34374991.0,
      "step": 14990
    },
    {
      "entropy": 5.2597309112548825,
      "epoch": 1.4404418828049952,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004798483324223533,
      "loss": 4.8647,
      "mean_token_accuracy": 0.23170171082019805,
      "num_tokens": 34386218.0,
      "step": 14995
    },
    {
      "entropy": 5.311018371582032,
      "epoch": 1.440922190201729,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004798341487309932,
      "loss": 4.8839,
      "mean_token_accuracy": 0.228623166680336,
      "num_tokens": 34396287.0,
      "step": 15000
    },
    {
      "epoch": 1.440922190201729,
      "eval_entropy": 5.0503300785844,
      "eval_loss": 4.936728477478027,
      "eval_mean_token_accuracy": 0.23430621224854672,
      "eval_num_tokens": 34396287.0,
      "eval_runtime": 20.6632,
      "eval_samples_per_second": 1588.091,
      "eval_steps_per_second": 198.517,
      "step": 15000
    }
  ],
  "logging_steps": 5,
  "max_steps": 104090,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7796586489845760.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}